SETTLING-TIME IMPROVEMENT IN GLOBAL CONVERGENCE LAGRANGIAN NETWORKS

Leonardo Acho

In this brief, a modification of Lagrangian networks given in (X. Youshen, 2003) is presented. This modification improves the settling time of the convergence of Lagrangian networks to a stationary point; which is the optimal solution to the nonlinear convex programming problem with linear equality constraints. This is important because, in many real-time applications where Lagrangian networks are used to find an optimal solution, such as in signal and image processing, this settling time is interpreted as the processing time. Simulation results applied to a quadratic optimization problem show that settling time is improved from about to 2000 to 20 seconds. Lyapunov theory was used to obtain our main result.

ADAPTIVE FUZZY SLIDING MODE CONTROLLER FOR THE SNORKEL UNDERWATER VEHICLE

Eduardo Sebastián and Miguel Ángel Sotelo Vázquez

This paper describes a control system for the kinematic variables of an underwater vehicle. Control of underwater vehicles is not simple, mainly due to the nonlinear, coupled and unknown character of system equations and dynamics. The proposed methodology makes use of a pioneer algorithm implemented for the first time in an underwater vehicle, and it is based on the fusion of a sliding mode controller and an adaptive fuzzy system, including advantages of both systems and relaxing the required knowledge of vehicle model.

ROBUST FUZZY CONTROLLER DESIGN FOR UNCERTAIN DESCRIPTOR MARKOVIAN JUMP SYSTEMS

Wudhichai Assawinchaichote and Sing Kiong Nguang

This paper examines the problem of designing a robust H_infty state-feedback controller for a class of uncertain nonlinear descriptor Markovian jump systems described by a Takagi-Sugeno (TS) fuzzy model with Markovian jumps. Based on a linear matrix inequality (LMI) approach, LMI-based sufficient conditions for the uncertain nonlinear descriptor Markovian jump systems to have an H_infty performance are derived. The proposed approach does not involve the separation of states into slow and fast ones and it can be applied not only to standard, but also to nonstandard nonlinear descriptor systems. A numerical example is provided to illustrate the design developed in this paper.

LOOKING FOR MASCONTROL: A MULTIAGENT SYSTEM FOR IDENTIFICATION AND CONTROL

E. J. González, Alberto Hamilton, L. Moreno, R. L. Marichal, J.A. Méndez and Vanessa Muñoz

In this paper, MASCONTROL, a multiagent system (MAS) for system identification and process control, is presented. This MAS implements a self-tuning regulator (STR) scheme. In this work, an Ontology Agent (OA) is included, using DAML+OIL as ontology language. From their experience, the authors consider this architecture highly useful for identification and control processes.

DEFECTIVE METAL END DETECTION WITH A FUZZY SYSTEM

Perfecto Mariño Espiñeira, Vicente Pastoriza Santos, Miguel Santamaría Sánchez and Emilio Martínez Expósito

The authors have been involved in developing an automated inspection system, based on machine vision, to improve the repair coating quality control (RCQ control) in can ends of metal containers for fish food. The RCQ of each end is assesed estimating its average repair coating quality (ARCQ). In this work we present a fuzzy model building to make the acceptance/rejection decision for each can end from the information obtained by the vision system. In addition it is interesting to note that such model could be interpreted and supplemented by process operators. In order to achieve such aims, we use a fuzzy model due to its ability to favour the interpretability for many applications. Firstly, the easy open can end manufacturing process, and the current, conventional method for quality control of easy open can end repair coating, are described. Then, we show the machine vision system operations. After that, the fuzzy modeling, results obtained and their discussion are presented. Finally, concluding remarks are stated.

A HIERARCHICAL FUZZY-NEURAL MULTI-MODEL: An application for a mechanical system with friccion identification and control

Ieroham Baruch, Jose Luis Olivares and Federico Thomas

A Recurrent Trainable Neural Network (RTNN) with a two layer canonical architecture and a dynamic Backpropagation learning method are applied for identification and control of complex nonlinear mechanical plants. The paper uses a Fuzzy-Neural Hierarchical Multi-Model (FNHMM), which merge the fuzzy model flexibility with the learning abilities of the RNNs. The paper proposes the application of two control schemes, which are: a trajectory tracking control by an inverse FNHMM and a direct adaptive control, using the states issued by the identification FNHMM. The proposed control methods are applied for a mechanical plant with friction system control, where the obtained comparative results show that the control using FNHMM outperforms the fuzzy and the neural control itself.

A FAST TABU SEARCH ALGORITHM FOR FLOW SHOP PROBLEM WITH BLOCKING

Jozef Grabowski and Jaroslaw Pempera

This paper develops a fast tabu search algorithm to minimize makespan in a flow shop problem with blocking. We present a fast heuristic algorithm based on tabu search approach. In the algorithm the multimoves are used that consist in performing several moves simultaneously in a single iteration of algorithm and guide the search process to more promising areas of the solutions space, where good solutions can be found. It allow us to accelerate the convergence of the algorithm. Besides, in the algorithm a dynamic tabu list is used that assists additionally to avoid trapped at a local optimum. The proposed algorithm is empirically evaluated and found to be relatively more effective in finding better solutions in a much shorter time.

FUZZY DIAGNOSIS MODULE BASED ON INTERVAL FUZZY LOGIC: OIL ANALYSIS APPLICATION

Antonio Sala, Bernardo Tormos, Vicente Macián and Emilio Royo

This paper presents the basic characteristics of a prototype fuzzy expert system for condition monitoring applications, in particular, oil analysis in Diesel engines. The system allows for reasoning under absent or imprecise measurements, providing with an interval-valued diagnostic of the suspected severity of a particular fault. A set of so-called metarules complements the basic fault dictionary for fine tuning, allowing extra functionality.

DISCRETE–TIME FREE AND FIXED END-POINT OPTIMAL CONTROL PROBLEM

Corneliu Botan and Florin Ostafi

A comparison between the fixed and free end-point discrete time linear quadratic optimal problem is performed. Symmetrical algorithms for both problems are proposed. These algorithms can be easier implemented by comparison with classical procedures. Simulation results are presented.

OPTIMIZED FUZZY SCHEDULING OF MANUFACTURING SYSTEMS

Nikolaos Tsourveloudis, Lefteris Doitsidis and Stratos Ioannidis

In this paper an Evolutionary Algorithm (EA) strategy for the optimization of generic Work-In-Process (WIP) scheduling fuzzy controllers is presented. The EA strategy is used to tune a set of fuzzy control modules that are used for distributed and supervisory WIP scheduling. The distributed controllers objective is to control the rate in each production stage in a way that satisfies the demand for final products while reducing WIP within the production system. The EA identifies those set of parameters for which the fuzzy controller performs optimal with respect to WIP and backlog minimization. The proposed EA strategy is compared with known heuristically tuned distributed and supervised fuzzy control approaches. Extensive simulation results show that the EA strategy significantly improves system’s performance.

MODEL PREDICTIVE CONTROL FOR DISTRIBUTED PARAMETER SYSTEMS USING RBF NEURAL NETWORKS

Eleni Aggelogiannaki and Haralambos Sarimveis

A new approach for the identification and control of distributed parameter systems is presented in this paper. A radial basis neural network is used to model the distribution of the system output variables over space and time. The neural network model is then used for synthesizing a non linear model predictive control configuration. The resulting framework is particular useful for control problems that pose constraints on the controlled variables over space. The proposed scheme is demonstrated through a tubular reactor, where the concentration and the temperature distributions are controlled using the wall temperature as the manipulated variable. The results illustrate the efficiency of the proposed methodology.

EMBEDDED ROBOTIC CONTROL TECHNOLOGIES AND ITS APPLICATIONS IN AUTOMATED PROGRAMMERS

Ganwen Zeng and Kelly Hirsch

The paper presents a synthesis of the problematic and actual solutions to the implementation of robotic programmer control functionality using DSP controllers. Considerable technology shift occurred during the recently decade in device programming industry. The advent of high performance DSP motion controllers opens new possibilities for the development of high performance distributed intelligence device-programming automation systems. The idea of implementing a unique, flexible robotic motion control structure can significantly improve controllability of the robotic programming systems. High-level motion command languages are used to setup and to control the robotic motors. A Fuzzy control algorithm has been introduced to guarantee the motion control performance in an automated programmer.

MODELLING HYBRID CONTROL SYSTEMS WITH BEHAVIOUR NETWORKS

Pierangelo Dell'Acqua, Anna Lombardi and Luís Moniz Pereira

We present an approach to model adaptive, dynamic hybrid control systems based on behaviour networks. We extend and modify the approach to behaviour networks with integrity constraints, non-ground rules, internal actions, and modules to make it self-adaptive and dynamic. The proposed approach is general, reconfigurable, robust, and suitable for environments that are dynamic and too complex to be entirely predictable, the controlling system having limited computational and time resources.

LQG CONTROL UNDER AMPLITUDE AND VARIANCE CONSTRAINTS

A. Królikowski, D. Horla and T. Kubiak

In this paper, the amplitude and variance-constrained LQG control is considered for a plant given by discrete-time ARMAX model. The minimization of constrained quadratic cost is approached by Kalman filter, approximation of the probability density function (pdf) of the state by the Gaussian one and by by tuning of the Lagrange multiplier. The obtained optimization algorithm is simulated for second-order stable plant model and different constraints.

ONTOLOGY FOR INTEGRATING HETEROGENEOUS TOOLS FOR SUPERVISION, FAULT DETECTION AND DIAGNOSIS

Beatriz López, Joaquim Meléndez and Silvia Suárez

The Distributed Supervision Systems that have been used extensively for the last fifteen years in the process industry are now evolving towards higher level solutions based on better connections between applications and processes that assure that data flows from the process to manage boards. Knowledge sharing seems to be a key issue in integrating these heterogeneous systems. In this paper we present an ontology as a first step to achieving semantic interoperability. The ontology has been conceived within the context of a complex integration problem, in which heterogeneous toolboxes cooperate to deal with several supervision, fault detection and diagnostic tasks for chemical processes. Regarding the current trends in ontology research, our proposal is consistent with top-level ontologies, as these kinds of ontologies seem to overcome the ontology integration problem. We describe a preliminary version of the ontology. The conceptualisation of control variables, system behaviour, supervision tasks, models and system properties is given. All attributes and relationships between each concept has been deployed. The ontology has been developed using Protete2000.

USE OF THE COG REPRESENTATION TO CONTROL A ROBOT WITH ACCELERATION FEEDBACK

Frédéric Colas, Eric Dumetz, Pierre-Jean Barre and Jean-Yves Dieulot

A controller using acceleration feedback has been applied to a flexible robot for which the position and velocity of the load are not measured. It is shown by using the Causal Ordering Graph (COG), that the motor can be controlled by using acceleration feedback and that it allows an exact tracking of the motor position, irrespective of the non-linear flexibilities of the axis and of the measurement disturbances. This easy-to-tune algorithm, which main control parameters are the modal masses of the motor and load part and only consists of a positive acceleration feedback plus a PD controller, has been validated on an industrial 3-axis robot.

FUZZY ADAPTIVE CONTROLLER FOR A SYNCHRONOUS MACHINE

Gregorio Drayer and Miguel Strefezza

This paper presents the comparison of applying an adaptive fuzzy controller with and without a variable structure controller (VSC) for a synchronous machine. A simplified linear model of the synchronous machine connected to an infinite bus with constant impedance is used. The multivariable system was previously decoupled to make easier the application of the control schemes. To control the system, an adaptive Fuzzy PD controller is proposed and it acts both on the load variable and on the voltage variable. Then, a Fuzzy Adaptive System is designed to act over the Fuzzy controller. After this, the VSC theory is applied to the Adaptive Controller to compare both strategies. Simulation results using these two control schemes are presented. With these proposed actions, the results show a better transitory response of the system when compared with the system response using classical control.

METHOD TO IMPROVE THE TRANSPARENCY OF NEUROFUZZY SYSTEMS

J. A. Domínguez-López

Neurofuzzy systems have been widely applied to a diverse range of applications because their robust operation and network transparency. A neurofuzzy system is specified by a set of rules with confidences. The use of rule confidences rather than a weight vector allows the model to be represented as a set of transparent fuzzy rules. Nevertheless, as knowledge base systems, neurofuzzy systems suffer from the curse of dimensionality i.e., exponential increase in the demand of resources and in the number of rules. Accordingly, the interpretability of the final model can be lost. Consequently, it is desired to have a simple rule-base to ensure transparency and implementation efficiency. After training, a rule can have several non-zero confidences. The more number of non-zero confidences, the less transparent the final model becomes. Therefore, it is elemental to reduce the number of non-zero confidences. To achieve this, the proposed algorithm search for (a maximum of) twon on-zero confidences which give the same result. Thus, the system can keep its complexity with a better transparency. The proposed methodology is tested in a practical control problem to illustrate its effectiveness.

GENETIC AND ELLIPSOID ALGORITHMS FOR NONLINEAR PREDICTIVE CONTROL

Kaouther Laabidi, Faouzi Bouani and Mekki Ksouri

This paper deals with the constrained predictive control of nonlinear systems. The Artificial Neural Networks (ANN) are used as a process model. The control law is derived by minimizing a non convex criterion. The optimization problem is solved using Ellipsoid and genetic algorithms. The structure and operators of the combining two algorithms have been specifically developed for control design problem. Simulation results are presented to illustrate the performance of the proposed predictive controller.

BIOPRODUCTS DRYING OPTIMAL CONTROL IN OSCILLATING REGIMES

Renat Sadykov, Dmitry Antropov and Rauf Kafiatullin

On the basis of the developed approaches and mathematical model (MM) of the bioactive products drying block is carried out the optimization problem of the equipment choice and its operation modes in view of deleted binary mixture an ethanol - water composition changes. The analysis of the problem with engaging of the Pontryagin’s maximum principle has revealed optimal control structure. There is developed the automated control system of drying installation with firmware, based on modern microprocessor technique. The guidelines on an drying processes intensification, worked out on the basis of the internal and external interconnected heatmasstransfer research, and the process optimal control considerably raise productivity of drying aggregates, reduce fuel and power expenditures.

SELF-LEARNING DISTURBANCE COMPENSATION FOR ACTIVE SUSPENSION SYSTEMS

Eckehard Münch, Henner Vöcking and Thorsten Hestermeyer

Ride comfort and safety of vehicles can be increased by active suspension systems. A problem is the detection of disturbances which can generally not be measured until they impact the chassis. Provided guidance and disturbance are known in advance, a controller can use this information to achieve considerably improved behavior. This paper presents an approach in which railway vehicles coupled in a network, in repeated runs over the same track section, learn a disturbance compensation that can almost entirely compensate for stationary disturbances, i.e., disturbances that occur at the same spot in equal measure. Here information on the respective track section is sampled, stored locally at the track, and retrieved by the succeeding vehicle which will use them for an improved compensation for the occurring disturbances and again store information there. This iterative procedure results in an optimal compensation. The algorithm is described and criteria for its design are derived from digital control theory. The procedure was implemented on a testbed for a semi-vehicle with three degrees of freedom. The results of the measurements are displayed and evaluated in this paper.

STABLE REPETITIVE CONTROL BY FREQUENCY ALIASING

James D. Ratcliffe, Paul L. Lewin, Eric Rogers, Jari J. Hätönen, Thomas J. Harte and David H. Owens

A filtering technique based on frequency aliasing which was initially developed for Iterative Learning Control is modified so that it can be implemented in real-time and is suitable for Repetitive Control. The aliasing technique is experimentally verified on a gantry robot facility, which manipulates payloads from a dispenser onto a constant velocity conveyor. A parallel arrangement consisting of a three-term feedback controller and a simple structure repetitive controller is used to control the robot. Without the aliasing technique, the combined control system becomes unstable very rapidly. In contrast, when the aliasing technique is applied, 1000 repetitions are successfully completed and no indications of impending instability can be seen.

MICROSILICON LUMINOUS FLUX SWITCH CONTROLLED BY MEANS OF MAGNETIC FIELD

J. Gołębiowski, T. Prohuń

The construction of a silicon beam which is used as a optical switch was presented. The investigated beam consists of three layers: on the silicon base the iron layer is put and it is followed by the aluminium layer. The change of the external magnetic field intensity causes the beam end displacement as well as the change of the luminous flux reflection angle. The influence of the magnetic transducer parameters as well as the field intensity on the luminous flux reflection angle are analysed. The optical system which is steered by the magnetic field was described.

GA BASED DATA FUSION APPROACH IN AN INTELLIGENT INTEGRATED GPS/INS SYSTEM

Ali Asadian, Behzad Moshiri, Ali Khaki Sedigh, and Caro Lucas

A new concept regarding to the GPS/INS integration, based on artificial intelligence here is presented. Most integrated inertial navigation systems (INS) and global positioning systems (GPS) have been implemented using the Kalman filtering technique with its drawbacks related to the need for predefined INS error model and observability of at least four satellites. Most recently, an INS/GPS integration method using a hybrid adaptive network based fuzzy inference system (ANFIS) has been proposed in leterature. The advantage of the ANFIS over other classical filtering algorithms is its ability to deal with noise in the input data in dynamic environments. During the availability of GPS signal, the ANFIS is trained to map the error between the GPS and the INS. Then it will be used to predict the error of the INS position components during GPS signal blockage. As ANFIS will be employed in real time applications, the change in the system parameters (e.g., the number of membership functions, the step size, and step increase and decrease rates) to achieve the minimum training error during each time period is automated. This paper introduces a genetic optimization algorithm that is used to update the ANFIS parameters with the INS/GPS error function used as the objective function to be minimized. The results demonstrate the advantages of the genetically optimized ANFIS for INS/GPS Integration in comparison with conventional ANFIS specially in the cases when facing satellites’ outages. Coping with this problem plays an important role in assessment of the fusion approach in land navigation.

A SCHEDULING TECHNIQUE OF PLANS WITH PROBABILITY AND TEMPORAL CONSTRAINTS

Bassam Baki and Maroua Bouzid

The paper describes a constraint programming approach for generating partially ordered plans with durative actions and probabilities. We propose a planner that generates a plan represented in the form of a set of tasks. Each task has a set of temporal constraints, a set of probabilities and a set of constant costs. All tasks form an acyclic AND/OR Graph in which our planner will find a plan formed by a set of tasks chosen to be executed in order to achieve a goal under specified constraints. This paper describes one approach to deal with a problem that has paid a little attention of planing community. This problem is to combine temporal and probabilistic planning.

INTEGRATED FEED-FORWARD ARTIFICIAL NEURAL NETWORKS SYSTEM FOR MACHINES TOOLS SELECTION

Romdhane Ben Khalifa, Noureddine Ben Yahia and Ali Zghal

We propose in this paper an integration module of the automatic choice of the machine tools in the environment of the systems CAD/CAM, which consisted in the two neuronal systems NN1 and NN2; NN1 allows the automatic choice of machining machines. NN2 makes it possible to choose cutting tools for machining features. In this work, we worked out two complementary parts for the integration of the automatic choice of machine tools. Firstly we developed a neuronal system for selection of machine tools classes. Secondly, one created an interface of integration of neuronal system which exploits the machining features geometrical data to be carried out by the programming Visual Basic

A HYBRID DECISION SUPPORT SYSTEM - The joint use of Simulation, Coloured Petri Nets and Expert System

Fabiano A. Hennemann, Ricardo J. Rabelo, José E. R. Cury, José V. Canto dos Santos and Arthur T. Gómez

This works aim to propose a Hybrid Decision Support System (HDSS), based in Simulation and Coloured Petri Nets as modelling techniques of manufacture processes, and an Expert System to assist in its use. The HDSS provides a friendly interface for the user that, after selecting input parameters, gets as answer a series of data about the manufacturing process that will assist in the evaluation of its performance. To validate the proposal, some particular scenes have been tested, with objective to elaborate a set of proposals for improving the performance of productive systems, evaluating the impacts from the change on model parameters and providing a better understanding about the systems considered. The HDSS makes it possible for managers, without knowledge of modelling techniques, to manipulate data and to interact with the developed model. The developed prototype was made generic for applying on general manufacturing processes, so that it is possible to use it for any industrial plant, provided that the input parameters of the model are adequately fitted, using the data input interface of the system.

APPLICATION OF DE STRATEGY AND NEURAL NETWORK - In position control of a flexible servohydraulic system

Hassan Yousefi and Heikki Handroos

One of the most promising novel evolutionary algorithms is the Differential Evolution (DE) algorithm for solving global optimization problems with continuous parameters. In this article the Differential Evolution algorithm is proposed for handling nonlinear constraint functions to find the best initial weights of neural networks. The highly non-linear behaviour of servo-hydraulic systems makes them idea subjects for applying different types of sophisticated controllers. The aim of this paper is position control of a flexible servo-hydraulic system by using back propagation algorithm. The poor performance of initial training of back propagation motivated to apply the DE algorithm to find the initial weights with global minima. This study is concerned with a second order model reference adaptive position control of a servo-hydraulic system using two artificial neural networks. One neural network as an acceleration feedback and another one as a gain scheduling of a proportional controller are proposed. The results suggest that if the numbers of hidden layers and neurons as well as the initial weights of neural networks are chosen well, they improve all performance evaluation criteria in hydraulic systems.

AN EXPLORATION MEASURE OF THE DIVERSITY VARIATION IN GENETIC ALGORITHMS

George Papakostas and Yiannis Boutalis

In this paper, a novel measure of the population diversity of a Genetic Algorithm (GA) is presented.Chromosomes diversity plays a major role for the successfully operation of a GA, since it describes the number of the different candidate solutions that the algorithm evaluates, in order to find the optimal one, in respect to a performance index, called objective function. In a well defined algorithm, the diversity of the current population should be measurable, in order to estimate the performance of the algorithm. The resulted observation, that is, the measuring of the diversity, can then be used to real-time adjust the factors that determine the chromosomes variety (Pc, Pm), during the execution of the GA. It is shown, that a simple chromosomes clustering into the search space, by using the well known k-means algorithm, can give a useful picture of the population’s distribution. Thus, by translating the problem of finding the best solution to a GA-based problem into an iterative clustering process, and by using the scatter matrices (Sw, Sb), which describe completely the candidate’s solutions topology, one could define a novel formula that gives the population diversity of the algorithm.

A NOVEL REPRESENTATION AND ALGORITHMS FOR (QUASI) STABLE MARRIAGES

B. Y. Zavidovique, N. Suvonvorn and Guna S. Seetharaman

In this paper, we propose "stable marriages" algorithms based on a novel representation called "marriage table". After explaining how properties as global satisfaction, sex equality and stability show in the representation, we define 3 algorithms corresponding to 3 different scans of the "marriage table" to meet progressively all constraints. The performance is evaluated in front of the population size for 200 instances in each case. That supports qualitative statistic analysis. Two matching examples in image processing are displayed for illustration.

KNOWLEDGE REPRESENTATION APPROACH TO CLOSED LOOP CONTROL SYSTEM - A TANK SYSTEM CASE-STUDY

Luís Rato, Irene Pimenta Rodrigues and Rui Gomes

Control engineering problems are dealt within a plethora of methods and approaches depending on the a priori knowledge, the description of the process to control, and the main control goal. Classical control theory is mainly based on properties of numerical models. This paper presents an approach that applies to a class of processes described by numerical and logical relations using inference and a knowledge base system. To attain this goal an ontology for control systems is constructed. The work presented in this paper is based in a three tank system benchmark.

D3G2A: A DYNAMIC DISTRIBUTED DOUBLE GUIDED GENETIC ALGORITHM FOR THE CASE OF THE PROCESSORS CONFIGURATION PROBLEM

BOUAMAMA Sadok and Khaled GHEDIRA

Within the framework of Constraint satisfaction and optimization problem (CSOP), we introduce a new optimization distributed method based on Genetic Algorithms (GA). This method consists of agents dynamically created and cooperating in order to solve the problem. Each agent performs its own GA on its own sub-population. This GA is sometimes random and sometimes guided by both the template concept and by the Min-conflict-heuristic. In addition with guidance, our approach is based on NEO-DARWINISM theory and on the nature laws. In fact, by reference to their specificity the new algorithm will let the agents able to count their own GA parameters. In order to show D3G2A advantages, experimental comparison with GGA is provided by their application on the Large processors configuration Problem.

OPTIMIZATION IN RAILWAY SCHEDULING

M. A. Salido, M. Abril, F. Barber, L. Ingolotti, A. Lova and P. Tormos

Train scheduling has been a significant issue in the railway industry. Over the last few years, numerous approaches and tools have been developed to aid in the management of railway infrastructure. In this paper, we present two filtering techniques for a constraint-based train scheduling tool, which is a project in collaboration with the National Network of Spanish Railways (RENFE), Spain. We formulate train scheduling as constraint optimization problems. Two filtering techniques are developed to speed up and direct the search towards suboptimal solutions in periodic train scheduling problems. The feasibility of our problem-oriented techniques are confirmed with experimentation using real-life data. The results show that these techniques enables MIP solvers such as LINGO and ILOG Concert Technology (CPLEX) to terminate earlier with good solutions.

DERIVING BEHAVIOR FROM GOAL STRUCTURE FOR THE INTELLIGENT CONTROL OF PHYSICAL SYSTEMS

Richard Dapoigny, Patrick Barlatier, Eric Benoit and Laurent Foulloy

Given a physical system described by a structural decomposition together with additional constraints, a major task in Artificial Intelligence concerns the automatic identification of the system behavior. We will show in the present paper how concepts and techniques from different AI disciplines help solve this task in the case of the intelligent control of engineering systems. Following generative approaches grounded in Qualitative Physics, we derive behavioral specifications from structural and equational information input by the user in the context of the intelligent control of physical systems. The behavioral specifications stem from a teleological representation based on goal structures which are composed of three primitive concepts, i.e. physical entities, physical roles and actions. An ontological representation of goals extracted from user inputs facilitates both local and distributed reasoning. The causal reasoning process generates inferences of possible behaviors from the ontological representation of intended goals. This process relies on an Event Calculus approach. An application example focussing on the control of an irrigation channel illustrates the behavioral identification process.

FEASIBLE CONTROL OF COMPLEX SYSTEMS USING AUTOMATIC LEARNING

Alejandro Agostini and Enric Celaya

Robotics applications often involve dealing with complex dynamic systems. In these cases coping with control requirements with conventional techniques is hard to achieve and a big effort has to be done in the design and tuning of the control system. An alternative to conventional control techniques is the use of automatic learning systems that could learn control policies automatically, by means of the experience. But the amount of experience required in complex problems is intractable unless some generalization is performed. Many learning techniques have been proposed to deal with this challenge but the applicability of them in a complex control task is still difficult because of their bad learning convergence or insufficient generalization. In this work a new learning technique, that exploits a kind of generalization called categorization, is used in a complex control task. The results obtained show that it is possible to learn, in short time and with good convergence, a control policy that outperforms a classical PID control tuned for the specific task of controlling a manipulator with high inertia and variable load.

MULTIOBJECTIVE OPTIMAL DESIGN OF STRUCTURE AND CONTROL OF A CONTINUOUSLY VARIABLE TRANSMISSION

Jaime Alvarez-Gallegos, Carlos A. Cruz-Villar and Edgar A. Portilla-Flores

An approach to solve the mechatronic design problem is to formulate the problem as a multiobjective dynamic optimization problem (MDOP), where kinematic and dynamic models of the mechanical structure and the dynamic model of the controller are considered besides a set of constraints and a performance criteria. This design methodology can provide a set of optimal mechanical and controller parameters so that the desired dynamic behavior and the performance criteria are satisfied. In this paper a MDOP is proposed and applied to a continuously variable transmission (CVT). Performance criteria are the mechanical efficiency and the minimal controller energy. The goal attainment method and a sequential approach are used to solve the MDOP.

CONTRIBUTORS TO A SIGNAL FROM AN ARTIFICIAL CONTRAST

Jing Hu, George Runger and Eugene Tuv

Data from a process or system is often monitored in order to detect unusual events and this task is required in many disciplines. A decision rule can be learned to detect anomalies from the normal operating environment when neither the normal operations nor the anomalies to be detected are pre-specified. This is accomplished through artificial data that transforms the problem to one of supervised learning. However, when a large collection of variables are monitored, not all react to the anomaly detected by the decision rule. It is important to interrogate a signal to determine the variables that are most relevant to or most contribute to the signal in order to improve and facilitate the actions to signal. Metrics are presented that can be used determine contributors to a signal developed through an artificial contrast that are conceptually simple. The metrics are shown to be related to traditional tools for normally distributed data and their efficacy is shown on simulated and actual data.

JAVA BASED TOOLBOX FOR LINEAR REPETITIVE PROCESSES

J. Gramacki, A. Gramacki, K. Gałkowski and E. Rogers

In the paper a Java based toolbox has been presented. It is used in teaching of a special case of nD systems - Linear Repetitive Processes (LRP). Its predecessor has been developed in the Matlab environment so to use it a Matlab licence is necessary. This restriction has been removed after making it available in the Internet network as a Java based program. Now a student may click a proper link on a web page and hence start an interactive work with a simulator of LRP. He or she may define a model as well as initial / boundary conditions, then simulate a process as a continuous or discrete case, analyze the results in graphical or numerical form, modify visualization parameters of the plots and finally print the results. In the paper an overview of the tool has been given.

ELECTRONIC AUTOMOTIVE REQUIREMENT DESIGN SPACE - A Bird’s Eye View of a Strategic Requirement Design Space Exploration

Liliana Díaz-Olavarrieta and David Báez-López

The purpose of this article is to make a holistic compilation of many different types of requirements for an automotive electronic communications / control network, and organize them into an easily reusable framework to help with the completeness / strategic consistency issues in the requirement specification process. The requirements framework proposed in this paper aims to answer the question: “What is the requirements design space for an automotive electronic communications network?”, and help in the completeness of the requirements specification through a holistic, multi-perspective, Bird’s Eye View. The main perspectives that will be examined in this requirements design space exploration are: a) Those derived from the “Nature of the User” perspective, b) Those derived from the “Nature of the Application” perspective: Distributed, Real time, Safety-Critical applications, and Resource Constraints requirements, c) Those derived from the “Nature of the Industry” competitive environment: Suppliers, Substitute Products / Technologies, Competitors, and Potential Entrants, the Company itself, its Clients and finally, d) Those derived from the “Nature of the Process Development” perspective, in particular, the component based development (CBD) process of Electronic Subsystem Design within Automotive Companies: component architecting, component assembly and component provisioning. The conceptual domain for the design of these specifications is the area of automotive electronic subsystems (known to be heterogeneous, distributed, real-time systems which in some cases have to implement safety-critical applications requiring fault-tolerant implementations), though the framework is in itself more generally applicable. The design and implementation of heterogeneous, real-time, distributed systems is a complex, knowledge intensive, problem. The design of embedded electronic distributed real-time systems for automotive applications, even more so. Indeed, the complexity comes not only from the electronics, but from all the non-electronic automotive parts which we currently view as “the car” – which interact with, constrain, and impact the electronic systems. The complexity can be handled by a variety of techniques, such as separation of concerns, layering and incremental development, iterative virtual modeling and simulation, and the use of validated automated design processes (such as the A, V models used in the automotive industry) to pass from one design/implementation phase to the next. Designs are generally validated against a set of specifications, both by testing of a system –both of its subsystems parts and their integration- (which is becoming more and more difficult in heterogeneous systems and later in the implementation process), or by following a design-rule constrained “refinement of specifications” within the Component Based Development paradigm that automotive manufacturers usually follow (due to outsourcing and supplier heterogeneity of mechanic, hydraulic and electronic subsystems). In order for the implementation to be correct, not only do the component subsystems have to be correct, the subsystem integration has to be correct and free of unintended interactions. The use of automated design tools starts after the specification or set of requirements for a system / subsystem have been decided upon. Thus, the issue of specification completeness, correctness, and consistency has to be dealt with, separately. The issue of correctness of the specification should be dealt with formal validation models. The issue of consistency can be handled through domain expert specification reviews. The lack of completeness of specifications is a “design specification flaw” which is difficult to detect, unless there is a reference model that one can use (i.e., we know that all states in a binary FSM must have 2 transitions defined –one for a “1” input and another for a “0”, and this knowledge can help to avoid specification flaws where some transition has not been defined). By analogy, if we do not have a higher level “requirements reference meta-model” (to tell us “all the requirements that you could ever think of specifying and don’t want to forget to consider”) we cannot know if the specification is complete. This paper proposes a novel “requirements meta-model” to explore the requirements design space.

EVOLUTIONARY COMPUTATION FOR DISCRETE AND CONTINUOUS TIME OPTIMAL CONTROL PROBLEMS

Yechiel Crispin

Nonlinear discrete time and continuous time optimal control problems with terminal constraints are solved using a new evolutionary approach which seeks the control history directly by evolutionary computation. Unlike methods that use the first order necessary conditions to determine the optimum, the main advantage of the present method is that it does not require the development of a Hamiltonian formulation and consequently, it eliminates the requirement to solve the adjoint problem which usually leads to a difficult two-point boundary value problem. The method is verified on two benchmark problems. The first problem is the discrete time velocity direction programming problem with the effects of gravity, thrust and drag and a terminal constraint on the final vertical position. The second problem is a continuous time optimal control problem in rocket dynamics, the Goddard's problem. The solutions of both problems compared favorably with published results based on gradient and nonlinear programming methods .

EFFICIENT LINEAR APPROXIMATIONS TO STOCHASTIC VEHICULAR COLLISION-AVOIDANCE PROBLEMS

Dmitri Dolgov and Ken Laberteaux

The key components of an intelligent vehicular collision-avoidance system are: sensing, evaluation, and decision making. We focus on the latter task of finding (approximately) optimal collision-avoidance control policies, which can naturally be modeled as a Markov decision process. Unfortunately, the exact classical MDP models and solution methods scale exponentially with the number of environment features, rendering them completely impractical for large-scale real-life domains. To address this, factored MDP representations and approximate solution algorithms have been proposed. In this work we apply approximate linear programming (ALP) to collision-avoidance problems, modeled as factored MDPs. Unlike the commonly-used primal ALP algorithms that approximate only the value function of the MDP, we investigate a composite approach that approximates both the objective function and the feasible region of the linear programs. Our empirical analysis demonstrates that we can obtain high-quality approximations to optimal control policies, while enjoying an exponential reduction in complexity (allowing us to solve problems whose complexity exceeds those solvable by standard MDP methods by tens of orders of magnitude).

ROBUST ILC DESIGN USING MÖBIUS TRANSFORMATIONS

C. T. Freeman, P. L. Lewin and E. Rogers

In this paper a general ILC algorithm is examined and it is found that the filters involved can be selected to satisfy frequency-wise uncertainty limits on the plant model. The probability of the plant model being at a given point in the uncertainty space is specified, and the filters are then chosen to maximise the convergence rate that can be expected in practice. The magnitude of the change in input over successive trials and the residual error have also been encorporated into the cost function. Experimental results are presented using a non-minimum phase test facility to show the effectiveness of the design method.

COOPERATIVE SELF-ORGANIZATION TO DESIGN ROBUST AND ADAPTIVE COLLECTIVES

Gauthier Picard and Marie-Pierre Gleizes

This paper aims at highlighting the benefits of using cooperation as the engine of adaptation and robustness for multi-agent systems. Our work is based on the AMAS (Adaptive Multi-Agent System) approach which considers cooperation as a self-organization mechanism to obtain adequate emergent global behaviors for systems coupled with complex and dynamic environments. A multi-robot resource transportation task illustrates the instantiation of a cooperative agent model equiped with both reactive and anticipative cooperation rules. Various experiments underline the relevance of this approach in dif?cult static or dynamic environments.

ON TEMPORAL DIFFERENCE ALGORITHMS FOR CONTINUOUS SYSTEMS

Alexandre Donzé

This article proposes a general, intuitive and rigorous framework for designing temporal differences algorithms to solve optimal control problems in continuous time and space. Within this framework, we derive a version of the classical TD($\lambda$) algorithm as well as a new TD algorithm which is similar, but designed to be more accurate and to converge as fast as TD($\lambda$) for the best values of $\lambda$, without the burden of finding these values.

REMOTE CONTROL FACILITIES OF WEB-BASED SURVEILLANCE SYSTEM FOR ELECTRIC POWER APPLIANCE AND NETWORK CAMERA

Yoshiro Imai, Yuichi Sugiue, Akira Andatsu, Daisuke Yamane, Hirofumi Kuwajima, Shin’ich Masuda

We have developed surveillance system, which had been organized with network cameras, an integrated web/mail server, mobile computing devices as GUI and remote control devices. Several kinds of devices can be used as our clients including, for example, high-performance cellular phone, which are equipped with Java virtual machine and web-browsing facilities. Our integrated server is designed to play intensive roles of web, e-mail and control services. It can obtain JPEG images from network cameras, process them, accumulate them into its database. It can also receive some types of requests from several kinds of clients, analyze them and perform already assigned services for monitoring and/or controlling. Almost all software of our surveillance system have been written in Java programming language, because of easy and powerful description of GUI as well as network programming. Users of our system can utilize remote monitoring and controlling anywhere and anytime, by means of mobile computing devices. Our integrated server can analyze the request from clients, generate the specific signals to subserver to switch several appliances as well as network cameras. Its subserver is able to control appliance power switching through the power line network, while its network cameras can be controlled by means of homing facilities of cameras themselves. With these facilities, remote control of electric power appliances and network cameras can be achieved by means of the commands from the above-mentioned integrated server.

A CONTROL SYSTEM USING BEHAVIOUR HIERARCHIES AND NEURO-FUZZY APPROACH

Dilek Arslan and Ferda N Alpaslan

In agent-based systems, especially in autonomous mobile robots, modelling the environment and its changes is a source of problems. It is not always possible to effectively model the uncertainty and the dynamic changes in complex, real-world domains. Control systems must be robust to changes and must be able to handle the uncertainties to overcome this problem. In this study, a reactive behaviour based agent control system is modelled and implemented. The control system is tested in a navigation task using an environment, which has randomly placed obstacles and a goal position to simulate an environment similar to an autonomous robot’s indoor environment. Then the control system was extended to control an agent in a multi-agent environment. The main motivation of this study is to design a control system, which is robust to errors and is easy to modify. Behaviour based approach with the advantages of fuzzy reasoning systems is used in the system

A NEW METHOD FOR WEIGHT UPDATING IN FUZZY COGNITIVE MAPS USING SYSTEM FEEDBACK

Theodore L. Kottas, Yiannis S. Boutalis and Manolis A. Christodoulou

Fuzzy Cognitive Maps (FCMs) have found many applications in social -fnancial -political problems. In this paper we propose a method of FCM operation, which can be used to represent and control any real system, including traditional electro-mechanical systems. In the proposed approach the FCM reaches its equilibrium point using direct feedback from the node values of the real system and the limitations imposed by the control objectives for the node values of the system. The experts’ knowledge, which is represented in the weights of the nodes’ interconnections, undergoes a continuous on-line adaptation based on feedback from the real system. An algorithm for weight updating is proposed, which is based on system feedback and which includes specially designed matrices that lead the FCM and consequently the real system associated with it in a balanced equilibrium state. The proposed methodology is tested by simulating the operation of a hydro-electric plant.

SPATIAL APPROACH IN RIVER BASIN MANAGEMENT USING DECISION MAKING STRATEGIES

Christian Menard

In this paper an approach towards a spatial decision support system is proposed for optimizing the management of river basins. All data from monitoring stations are collected and stored in a centralized database system. Since all measurement data are spatial and time related, spatial services fulfill the requirements in a decision making process best. A spatial decision support system approach is presented in which modeling is based on a network structure. This network can then be used to design and calibrate the underlying model. Spatial information can be obtained directly using GIS functionality.

STATIONARY FULLY PROBABILISTIC CONTROL DESIGN

Tatiana V. Guy and Miroslav Kárný

Stochastic control design chooses the controller that makes the closed-loop system behavior as close as possible to the desired one. The considered fully probabilistic design employs probabilistic description of both closed-loop and desired behaviors and uses Kullback-Leibler divergence as their proximity measure. An explicit minimiser provided by this design allows simpler approximation of analytically non-feasible cases. The existing formulations are oriented towards finite-horizon design and lead to the non-stationary optimal strategy. The paper provides infinite-horizon problem formulation and corresponding solution. This leads to a stationary strategy, which approximation is much easier.

CONTROL FOR ELECTRICAL NEUROMUSCULAR STIMULATOR USING FUZZY LOGIC - Trainning gait in paraplegics

Leonardo Rodrigues da Silva and Percy Nohama

This article presents a personal computer-based control system for an electrical stimulator using fuzzy logic. The input signal comes from a goniometer and the output is the stimulation level to be applied in the muscle of the patient. By this way, that control system is made for the therapist that just specifies the desired joint angle. The movement that the patient will execute can be mimicked from a person with normal movements, storing his or her joint’s angles during the execution of some task, and later reproducing it in the person without the voluntary movements. Such movements will be more proper of a human than a planned execution of a computational system, which the movement is structuralized by means of vectors, angles and times placed of supposed form.

MILITARY VEHICLE TYPE CLASSIFICATION - Intelligent Control Systems and Optimization

Jerzy Jackowski

This work presents the results of the measurement of the noise generated by vehicles differentiated in respect of the vehicle weight and structure. The analysis of registered acoustic signals was carried out on the basis of their frequency representation. Based on the Student difference test, a series of parameters of determined spectral signal power densities were examined for their usefulness for a differentiating feature vector. A process of qualifying a registered signal of a detected object to a proper class can be realized by various methods. Most often it is carried out on the basis of the object feature vector position against surfaces separating it from the vectors of other objects in the multidimensional space of features. Meeting the requirement of maximum classifier structure simplification, searching for the best separating plane was limited to the neuron network method based on the Rosenblatt perceptron education. Specification of measurement results indicates that there is a high probability of correct recognition of acoustic signals generated by the wheel and caterpillar vehicle motion.

REAL-TIME TIME-OPTIMAL CONTROL FOR A NONLINEAR CONTAINER CRANE USING A NEURAL NETWORK

T. J. J. van den Boom, J. B. Klaassens and R. Meiland

This paper considers time-optimal control for a container crane based on a Model Predictive Control approach. The model we use is nonlinear and it is planar, i.e. we only consider the swing (not the skew) and we take constraints on the input signal into consideration. Since the time required for the optimization makes time-optimal not suitable for fast systems and/or complex systems, such as the crane system we consider, we propose an off-line computation of the control law by using a neural network. After the neural network has been trained off-line, it can then be used in an on-line mode as a feedback control strategy.

INTELLIGENT MOBILE MULTI-ROBOTIC SYSTEMS: SOME CHALLENGES AND POSSIBLE SOLUTIONS

Flávio S. Corrêa da Silva, Renata Wassermann, Ana Cristina V. Melo, Leliane N. Barros and Marcelo Finger

Intelligent mobile multi-robotic systems (IMMRSs) are coordinated systems of autonomous mobile robots endowed with reasoning capabilities. This sort of systems requires the integrated application of a variety of state-of-the-art techniques developed within the realm of Artificial Intelligence, as well as instigates the further development of different specialisations of Artificial Intelligence. In the present article we examine some of these techniques and specialisations, discuss some specific challenges proposed to the field of Artificial Intelligence by IMMRSs, and suggest possible solutions to these challenges. In order to make our presentation more concrete, we employ throughout the article a specific example of IMMRS application, namely security surveillance of an empty building by a team of robots - the well known pursuit-evasion problem.

PEDESTRIAN RECOGNITION FOR INTELLIGENT TRANSPORTATION SYSTEMS

D. Fernández, I. Parra, M. A. Sotelo, L. M. Bergasa, P. Revenga, J. Nuevo and M. Ocaña

This paper describes a binocular vision-based pedestrian recognition System. The basic components of pedestrians are first located in the image and then combined with a SVM-based classifier. This poses the problem of pedestrian detection and recognition in real, cluttered road images. Candidate pedestrians are located using a subtractive clustering attention mechanism. A distributed learning approach is proposed in order to better deal with pedestrians variability, illumination conditions, partial occlusions and rotations. The performance of the pedestrian recognition system is enhanced by a multiframe validation process. By doing so, the detection rate is largely increased. A database containing hundreds of pedestrian examples extracted from real traffic images has been created for learning purposes. We present and discuss the results achieved up to date.

EXTENSION VERSUS BENDING FOR CONTINUUM ROBOTS

Robin McDonnell, George Grimes, Ian D. Walker and Carlos Carreras

In this paper, we analyze the capabilities of a novel class of continuous-backbone (“continuum”) robots. These robots are inspired by biological “trunks, and tentacles”. However, the capabilities of established continuum robot designs, which feature controlled bending but not extension, fall short of those of their biological counterparts. In this paper, we argue that the addition of controlled extension provides dual and complementary functionality, and correspondingly enhanced performance, in continuum robots. We present an interval-based analysis to show how the inclusion of controllable extension significantly enhances the workspace and capabilities of continuum robots.

COMPOSITIONAL ANALYSIS FOR REGULARITY, LIVENESS AND BOUNDEDNESS

Li Jiao

Desel introduced regular nets by the linear algebraic representation of nets [1]. Regularity is a sufficient condition for an ordinary net to be live and bounded. All the conditions checking the regularity of a net are decidable in polynomial time in the size of a net [2]. This paper proves that regularity, liveness and boundedness can be preserved after many compositional operations. This means that one designer can construct complex nets satisfying regularity, liveness and boundedness properties from simpler ones without forward analysis.

TOPOLOGICALLY ROBUST RECONSTRUCTION OF A 3D OBJECT WITH ORGANIZED MESHING

Junta Doi and Wataru Sato

This study proposes a topologically robust and noise-resistive reconstruction procedure that approximates a real 3D object. A geometric model with desired meshing is directly reconstructed based on a solid modeling approach. The radial distance of each scanning point from the axis of the cylindrical coordinates is measured using a laser triangulation sensor. The angular and vertical positions of the laser beam are two other coordinate values of the modeling. A face (mesh) array listing (topology), which defines the sampling point connectivity to form the mesh and the shape of the mesh, is assigned to meet the meshing The topologically stable and thus organized meshing, and hence, an accurate approximation is then accomplished. It is free from the noise-originated misconnection and shape ambiguity, which is unavoidable in the recent ICP (Iterative Closed Point) modeling. This proposal allows a versatile 3D shape processing and modification, for instance, for cultural heritage retrieval and virtual training.

A NEW ART GALLERY ALGORITHM FOR SENSOR LOCATION

Andrea Bottino and Aldo Laurentini

Locating sensors in2D can be often modeled as an Art Gallery problem. Unfortunately, this problem is NP-hard, and no finite algorithm, even exponential, is known for its solution. Algorithms able to closely approximate the optimal solution and computationally feasible in the worst case are unlikely to exist. However this is an important problem, and algorithms with “good” performance in practical cases are sorely needed. After reviewing the available algorithms, we propose a new sensors location incremental technique. The technique converges toward the optimal solution. It locally refines a starting approximation provided by an integer covering algorithm, where each edge is observed entirely by at least one sensor. A lower bound for the number of sensor, specific of the polygon considered, is used for halting the algorithm, and a set of rules are provided to simplify the problem.

AN OPTIMAL CONTROL SCHEME FOR A DRIVING SIMULATOR

Hatem Elloumi, Marc Bordier and Nadia Maïzi

Within the framework of driving simulation, control is a key issue to providing the driver realistic motion cues. Visual stimulus (virtual reality scene) and inertial stimulus (platform motion) induce a self-motion illusion. The challenge is to provide the driver with the sensations he would feel in real car maneuvering. This is an original control problem. Indeed, the first goal is not classical path tracking but fooling the driver awareness. Constrained workspace is the second issue classically addressed by motion cueing algorithms. The purpose of this paper is to extend the works of Telban and Cardullo on the optimal motion cueing algorithm. A nonlinear dynamical model of the robot is brought in. The actuator forces are directly included in the optimal control scheme. Consequently a better (global) optimization and an advanced parametrization of the control are achieved.

STATE TRANSFORMATION FOR EULER-LAGRANGE SYSTEMS

M. Mabrouk and J. C. Vivalda

The transformation of an Euler-Lagrange system into a state affine system in order to solve some interesting problem as the design of observer, the output tracking control, is considered in this paper. A necessary and a sufficient condition is given as well as a method to compute this transformation.

IMPROVEMENT OF THE DYNAMICS OF THE CONTINUOUS LINEAR SYSTEMS WITH CONSTRAINTS CONTROL

N.H. Mejhed, A. Hmamed and A. Benzaouia

A time varying control law is proposed for linear continuous-time systems with non Symmetrical constrained control. Necessary and sufficient conditions allowing us to obtain the largest non-symmetrical positively invariant polyhedral set with respect to (w.r.t) the system in the closed loop are given. The asymptotic stability of the origin is also guaranteed. The case of symmetrical constrained control is obtained as a particular case. The performances of our regulator with respect to the results of [3] are shown with the help of an example

PATH FOLLOWING IN UNKNOWN ENVIRONMENT FOR A CAR-LIKE MOBILE ROBOT

Niramon Ruangpayoongsak and Hubert Roth

The path following is the automatic control of the mobile robot movement along the specified path without human interference. The proposed path following applies for the robot navigation in unknown environments, where the robot has no preliminary information about obstacles. An innovative idea for the path following control is to integrate the basic path following control with the obstacle avoidance and the trajectory generation. The basic path following control is first implemented without obstacle detection. The robot receives the desired path, performs driving along the path, and stops at the destination. The obstacle avoidance is developed by wall following technique using on ultrasonic and infrared sensors. The trajectory generation is to generate the fittest trajectory between current position and the destination position after the robot is free from obstacles. These algorithms base on the car manoeuvring characteristics.

THERMAL SPRAYING ROBOT KINEMATICS AND LASER PATTERN CONTROL

Dermot Breen, Eugene Coyle and David Kennedy

The thermal spraying surface engineering industry relies on manual spraying and standard pre-programmed robotic systems. This research presents the completed geometric forward and inverse kinematics solution for a non standard articulated robotic manipulator which includes continuous 3600 axis rotation for waist, shoulder and elbow joints with a commercially available joint for tilt and pitch. The research also details the use of PTFE electroless nickel slip rings and brushes for providing delivery of power and data through the 3600 continuous rotation joints. The automatic analysis of distance and orientation measurement via a pattern producing laser and camera system are described which can be applied to the thermal spraying process for automatic feedback control of the robotic arm manipulator. The competed technical and simulation design will provide for the automatic application of advanced surface coatings to enhance wear, low friction and corrosion resistance properties to substrates via a thermal spraying process.

PREDICTIVE CONTROL FOR MODERN INDUSTRIAL ROBOTS - Algorithms and their applications

Květoslav Belda, Josef Böhm and Pavel Píša

Industrial robots comprise substantial parts of machine tools and manipulators in production lines. Their present development stagnates in their control. Traditional approaches, e.g. NC systems combined with PID/PSD structures, provide control of the tool drives as separate units only, but not solve the control from view of the whole machine system. On the other hand, in control theory, there are a lot of approaches, in which the information on tool dynamics and kinematic relations can be involved. The main contribution of this paper is to introduce various utilization and modifications (not only control tasks) of one such approach – model-base predictive control. The control is being developed for modern industrial robots based on parallel configurations. The modifications of predictive algorithm are substantiated by real laboratory experiments. The paper concerns with basic control design, possibilities of removing positional steady-state error. Quadraticaly-optimal trajectory planning is outlined in it.

TRACKING OF A UNICYCLE-TYPE MOBILE ROBOT USING INTEGRAL SLIDING MODE CONTROL

Michael Defoort, Thierry Floquet, Wifrid Perruquetti and Annemarie Kokosy

This paper deals with the tracking control for a dynamic model of a wheeled mobile robot in the presence of some perturbations. The control strategy is based on integral sliding mode. Simulations illustrate the results on the studied mobile robot.

MULTIPLE VIEW GEOMETRY ESTIMATION BASED ON FINITE-MULTIPLE EVOLUTIONARY AGENTS FOR MEDICAL IMAGES

Mingxing Hu, Karen McMenemy, Stuart Ferguson, Gordon Dodds and Baozong Yuan

In this paper we present a new method for the robust estimation of the trifocal sensor, from a series of medical images, using finite-multiple evolutionary agents. Each agent denotes a subset of matching points for parameter estimation, and the dataset of correspondences is considered as the environment in which the agents inhabit, evolve and execute some evolutionary behavior. Survival-of-finite-fitness rule is employed to keep the dramatic increase of new agents within limits, and reduce the chance of reproducing unfit ones. Experiments show that our approach performs better than the typical methods in terms of accuracy and speed, and is robust to noise and outliers, even when a large number of outliers are involved.

SIMULATING TELEROBOTICS BY CELLULAR TELEPHONY

Rodrigo Montúfar-Chaveznava, Crisóforo Paisano and José María Cañas

In this work we present a simulation system to control a mobile robot using a cell phone. We exploit the cellular telephony technology to communicate and command a robot provided with a modem in a simulation program. The system considers that telerobotics can be carried out using phone tones or sending text messages via SMS. Due the cellular telephony is expensive, the simulation system give us the possibility to experiment infinitely this novel way of telerobotics without expending economical resources. We design all components required for a basic cellular telephony system and we employ a robotics simulator for the Pioneer mobile robot expecting to translate the system primitives to real systems in a near future. The idea comes up due that cellular telephony has an enormous covering, and taking advantage of such situation, telerobotics could be performed in places where wireless networking and power sources are not available at all such as in the country.

OBSTACLE DETECTION BY STEREO VISION, INTRODUCING THE PQ METHOD

H. J. Andersen, K. Kirk, T. L. Dideriksen, C. Madsen and M. B. Holte

Safe, robust operation of an autonomous vehicle in cross-country environments relies on sensing of the surroundings. Thanks to the reduced cost of vision hardware, and increasing computational power, computer vision has become an attractive alternative for this task. This paper concentrates on the use of stereo vision for obstacles detection in cross-country environments where the ground surface can not be modeled as ramps, i.e. linear patches. Given a 3D reconstruction of the surrounding environment, obstacles are detected using the concept of compatible points. The concept classify points as obstacles if they fall within the volume of cone located with its apex at the point being evaluated. The cone may be adjusted adjusted according the physical parameters of the vehicle. The paper introduces a novel Projection and Quantification method that based on vehicle orientation rotates the 3D information onto an intuitive two dimensional surface plot and quantifies the information into bins adjusted to the specific application. In this way the search space for compatible points is significantly reduced. The new method is evaluated for a specific robotic application and the results are compared to previous results on a number of typical scenarios. Combined with an intuitive representation of obstacles in a two dimensional surface plot, the results indicate a significant reduction in the computational complexity for relevant scenarios.

A SWITCHING ALGORITHM FOR TRACKING EXTENDED TARGETS

Andreas Kräußling, Frank E. Schneider and Dennis Wildermuth

Tracking extended objects like humans in crowded environments is one of the challenges in mobile robotics. Several characteristics must be taken into consideration when evaluating the performance of such a tracking algorithm - e.g. accuracy, the need for computation time and the ability to deal with complex situations like crossing targets. In this paper two different algorithms for tracking extended targets are examined and compared by means of these criterions. One result is that none of the algorithms alone is a sufficient solution to the criterias. Therefore, a switching approach using both algorithms is introduced and tested on real data.

BROKEN BAR DETECTION IN INDUCTION MOTORS - Using non intrusive torque estimation techniques

Mario Eltabach and Ali Charara

One of the most important issues when implementing control and fault diagnosis systems for induction motor drives is obtaining accurate information about the state of certain motor electromagnetic signals such as stator flux and electromagnetic torque. This paper examines the detection of rotor imperfections through spectral analysis of the electromagnetic torque, computed by three stator flux estimators, and using only non-invasive sensors such as current and voltage sensors. The variable structure observer, the extended Luenberger observer (ELO) and extended Kalman filter (EKF) are used to estimate flux components without resorting to the use of intrusive speed sensors. The aim of this paper is to make a comparison and a classification between these approaches. Experimental results demonstrate the significant potential of these methods in detecting these types of faults

PATTERN RECOGNITION FEATURE AND IMAGE PROCESSING THEORY ON THE BASIS OF STOCHASTIC GEOMETRY

Nikolay G. Fedotov, Lyudmila A. Shulga, Alexander V. Moiseev and Andrey S. Kol’chugin

Application of stochastic geometry methods to pattern recognition is analysed. The paper is based on Trace-transformations of original images introduced by [1] into images on the Möbius band. Based on the new geometric transformation, a new approach towards the construction of features, independent of images’ motions or their linear transformations, is put forward. A prominent characteristics of the group of features under consideration is that we can represent each of them as a consecutive composition of three functionals. The paper considers the application of three-functional structure of recognition feature to image pre-processing. Feature can be invariant or sensitive to the affine transformation and linear deformation of objects depending of functionals selection. Thus sensitive features are suitable to determine the parameters of translation. It is an important task for robotics.

HIGHLY ACCURATE INTEGRATION OF TRACK MOTIONS

Michael Kleinkes, Angela Lilienthal and Werner Neddermeyer

According to the largeness of the workpieces in several industrial environments, a great number of industrial robots is placed on external track motions, so called 7th axes, as for automotive or aircraft industries. Flexible automation today requires absolute high accuracy. For example modern robotics deals with offline programming, copying programs between similar working cells, reflecting programs or image processing for 3D-pose estimation. All these tasks need absolute high accuracy and in fact, there have been many investigations for increasing the accuracy of single robots in the past few years. In contrast to that the use of track motions will dramatically increase the position error and badly influence the static behaviour of the robot system. The main reasons for these additional errors are the incorrect identi- fication of the main track direction and furthermore, very crucial, the non-linearities of the TCP (Tool Center Point) during the robots motion on the track. This article will introduce a new method of identification and mathematical integration of linear tracks. At first we present the method for measuring and generating profiles of single tracks by making use of the discrete fourier transformation (DFT) and cubic spline interpolation. Then a method for recalculating offline generated programs for real environments is presented, followed by a method for copying programs taking two profiles of track motions into consideration. Finally some measurement results are shown.

ON-LINE SUPERVISED ADJUSTMENT OF THE CORRECTING GAINS OF FRACTIONAL ORDER HOLDS

A. Bilbao-Guillerna, M. De la Sen and S. Alonso-Quesada

A discrete control using different possible discretization models of a continuous plant is presented. The different models of the scheme are obtained from a set of different discretizations of a continuous transfer function under a fractional-order-hold of correcting gain β. The objective is to design a supervisory scheme which is able to find the most appropriate value for the gain β in an intelligent design framework. A tracking performance index evaluates each possible discretization and the scheme chooses the one with the lowest value. Two different methods of adjusting this value are presented and discussed. The first one selects it among a fixed set of possible values, while in the second one the value of can be updated by adding or subtracting a small quantity. Simulations are presented to show the usefulness of the scheme.

ROBUST CONTROL OF INDUCTION MOTOR USING FAST OUTPUT SAMPLING TECHNIQUE

Alemayehu G.-E. Abera, B. Bandyopadhyay, S. Janardhanan and Vivek Agrawal

In this paper a design method based on robust fast output sampling technique is presented for the speed control of induction motor. The nonlinear model of induction motor model is linearized around various operating points to obtain the linear models. The input of the induction motor is the stator voltages and only the speed is considered as the output of the systems. A single controller is designed for these linear models. The nonlinear model of the induction motor is simulated with the proposed controller at these operating points. This method does not require the state of the system for feedback and is easily implementable.

MODELLING AND LQ-BACKSTEPPING CONTROL FOR A QUADROTOR

Sébastien Belin, Mathieu Carton and Fabien Macaire

Thanks to significant advances during the last decades in the miniaturized robotic area, many Unmanned Aerial Vehicle (UAV) projects were launched. Among them, the QuadriXflyer is an UAV quadrotor designed to evolve autonomously between waypoints given by an operator before flight. In this paper, we propose a modelling and a new hybrid control approach for the QuadriXflyer; a controller integrating the advantages of a Linear Quadratic (LQ) and those of a backstepping approach allowing to compensate the nonlinearities of the system. With this new approach, the gravity will be compensated directly without time delay. Robustness of the controller is then studied to ensure the stability of the quadrotor to exogenous (wind for example) and internal (noise on measurements, uncertainties on the inertia for example) perturbations.

AN OPEN OBJECT ORIENTED PATH PLANNING SYSTEM

Eleonora Fantini, Monica Reggiani and Stefano Caselli

The paper describes the ongoing development of a motion planning system whose aim is to ease the study and development of new planning strategies as well as the benchmarking and comparison of the existing ones. The system is implemented using open technologies and exploiting advanced object-oriented programming concepts. It efficiently integrates multiple planning strategies and collision detection algorithms and provides support for diverse geometric formats.

A UNIVERSAL MODULAR AUTONOMOUS ROBOT ARCHITECTURE

Wolfgang Ertel, Joachim Fessler and Nico Hochgeschwender

We present a universal modular robot architecture. A robot consists of the following intelligent modules: central control unit (CCU), drive, actuators, a vision unit and sensor input unit. Software and hardware of the robot fit into this structure. We define generic interface protocols between CCU and drive and between CCU and vision unit. If the robot has to solve a new application and is equipped with a different drive, new actuators and different sensors, only the program for the new application has to be loaded into the CCU. The interfaces to the drive, the vision unit and the other sensors are plug-and-play interfaces. The only constraint for the CCU-program is the set of commands for the actuators. Communication between the CCU and the vision unit uses ethernet. All the other components communicate via a CAN-bus. The domain of robot soccer serves as a first test case for this architecture.

SELF-KNOWLEDGE BASED ON THE ATOMIC CAPABILITIES CONCEPT - A Perspective to Achieve Sure Commitments among Physical Agents

Christian G. Quintero M., Josep Ll. de la Rosa and Josep Vehí

This paper presents an engineering perspective based on the atomic capabilities concept (AC2) to include control-oriented knowledge in the decisions making structure of physical agents (e.g. mobile robots). These agents operate in a real environment managing physical objects (e.g. their physical bodies) in coordinated tasks. AC2 guarantees an appropriate agents-oriented representation about the specifications of automatic controllers installed within the physical agents. This approach allows to each agent a reliable self-knowledge which concludes in achieving sure commitments and intelligent control in a cooperative system. Examples and conclusions are presented, emphasising the advantages of our proposal in the multi-agent system performance in physical environments.

SYNTHESIZING DETERMINISTIC CONTROLLERS IN SUPERVISORY CONTROL

Andreas Morgenstern and Klaus Schneider

Supervisory control theory for discrete event systems is based on finite state automata whose inputs are partitioned into controllable and uncontrollable events. Well-known algorithms used in the Ramadge-Wonham framework disable or enable controllable events such that it is finally possible to reach designated final states from every reachable state. However, as these algorithms compute the least restriction on controllable events, their result is usually a nondeterministic automaton that can not be directly implemented. For this reason, one distinguishes between supervisors (directly generated by supervisory control) and controllers that are further restrictions of supervisors to achieve determinism. Unfortunately, controllers that are generated from a supervisor may be blocking, even if the underlying discrete event system is nonblocking. In this paper, we give a modification of a supervisor synthesis algorithm that enables us to derive deterministic controllers. Moreover, we show that the algorithm is both correct and complete, i.e., that it generates a deterministic controller whenever one exists.

STEREO IMAGE BASED COLLISION PREVENTION USING THE CENSUS TRANSFORM AND THE SNOW CLASSIFIER

Christian Küblbeck, Roland Ach and Andreas Ernst

In this paper we show an approach for a stereo-camera based system on a moving roboter to avoid obstacles. We propose the ``census transformation'' for generating the feature for the correspondence search. We train two SNOW (spare network of winnovs)-classifiers, one for the decision wether to move straight forward or to evade and a second one for deciding whether to tunr left or right when evading. For training we use a sample set collected by manually moving around with the robot platform. We evaluate the performance of the whole recognition chain (feature generation and classification) using ROC-curves. Real world experiments show the mobile robot to safely avoid obstacles. Problems still arise when approaching steps or low obstacles due to limitations in the camera setup. We propose to solve this problem using a stereo camera system capable of pann and tilt movements.

REVERSIBILITY ENFORCEMENT FOR UNBOUNDED PETRI NETS

Hanife Apaydın Özkan and Aydın Aybar

In this paper, partially reversibility property and reversibility enforcement are studied for unbounded Petri nets. A method which tests partial reversibility, and also finds a bound vector guaranting reversibility for unbounded Petri nets is developed and an algorithm of the method is generated. Furthermore a controller design approach which enforces reversibility for unbounded Petri nets is introduced.

VERIFICATION OF TIMED CHI MODELS USING UPPAAL

E. M. Bortnik, D. A. van Beek, J. M. van de Mortel-Fronczak and J. E. Rooda

Due to increasing system complexity and growing competition and costs, powerful techniques are needed to design and analyze manufacturing systems. One of the most popular techniques to do performance analysis is simulation. However, simulation-based analysis cannot guarantee the correctness of a system. Our research focuses on examining other methods to make performance analysis and functional analysis, and combining the two. One of the approaches is to translate a simulation model that is used for performance analysis to a model written in an input language of an existing verification tool. The process algebraic language Chi is intended for modeling, simulation, verification and real-time control and has been used extensively to simulate large manufacturing systems. Uppaal is an integrated tool environment for modeling, validation and verification of real-time systems and has been applied successfully in case studies ranging from communication protocols to multimedia applications. In this article, we represent a translation scheme that is used to translate simulation models written in Chi language to Uppaal timed automata. As an example we apply the translation scheme to a model of a manufacturing system and show some of the properties that can be verified in Uppaal.

A STRATEGY FOR BUILDING TOPOLOGICAL MAPS THROUGH SCENE OBSERVATION

Roger Freitas, Mário Sarcinelli-Filho, Teodiano Bastos-Filho and José Santos-Victor

Mobile robots remain idle during significant amounts of time in many applications, while a new task is not assigned to it. In this paper, we propose a framework to use such periods of inactivity to observe the surrounding environment and learn information that can be used later on during navigation. Events like someone entering or leaving a room, someone approaching a printer to pick a document up, etc., convey important information about the observed space and the role played by the objects therein. Information implicitly present in the motion patterns people describe in a certain workspace is then explored, to allow the robot to infer a “meaningful” spatial description. Such spatial representation is not driven by abstract geometrical considerations but, rather, by the role or function associated to locations or objects (affordances) and learnt by observing people’s behaviour. Map building is thus bottom-up driven by the observation of human activity, and not simply a top-down oriented geometric construction.

ONLINE ESTIMATION OF SHIP STEERING DYNAMICS AND ITS APPICATIONS IN DESIGNING AN OPTIMAL AUTOPILOT

Minh-Duc LE and Hai-Nam Nguyen

Recursive Least Square (RLS) Algorithm applied to a Multivariate Auto-Regressive (MAR) process is used to estimate ship steering dynamics online. The estimation method is then linked to the Linear Quadratic (LQ) Algorithm to design an optimal autopilot for steering ships. The estimation method was applied to several ships and model ships and in all the cases the estimated parameters converged well. The design algorithm was used to construct a tracking system for course keeping and course changing maneuvers. Simulation results for the ships show the robustness of the estimation method and prove that the autopilot has very good performance.

COMPONENT RUNTIME SELF-ADAPTATION IN ROBOTICS

Daniel Hernandez, Antonio Dominguez, Oscar Deniz and Jorge Cabrera

Mobile robotic applications have to deal with limited resources and variable execution conditions that must be appropriately managed in order to keep an acceptable system behavior. This requires the implementation of runtime adaptation mechanisms that monitor continuously system state and module the resulting performance as a function of the available resources. As we consider that these adaptation mechanisms should be offered as a facility to robotic application programmers, we have integrated them inside CoolBOT, a component oriented framework for programming robotic systems. CoolBOT contributes to reduce the programming effort, promoting code reuse, while the adaptation scheme allows for more robust applications with an extended range of operation. In this paper we also present two demonstrators that outline the benefits of using the proposed approach in the development of real robotic applications.

LOCALIZATION FOR A CAR-LIKE MOBILE ROBOT USING NONLINEAR DYNAMIC MODEL

Niramon Ruangpayoongsak, Hubert Roth and Robert Mayr

The problem of localization is well known in mobile robotics. A solution is to use a model-based technique such as a kalman filter with multi sensor data fusion. For a car-like mobile robot, the nonlinear dynamic model is suitable for robot movement representation. This work presents the discrete extended kalman filter including a nonlinear dynamic model for the mobile robot localization. As inputs for the kalman filtering, gyroscope and compass sensors provide the relative and absolute yaw angles. The experiments are performed on several path types and the averages of the final position errors and the final heading errors are proposed.

EVOLUTIONARY LEARNING OF FUZZY RULES IN A MODIFIED CLASSIFIER SYSTEM FOR MOBILE AGENTS CONTROL

Eric Vallejo Rodríguez and Ginés Benet Gilabert

In this work we present the creation of a platform, along with an algorithm to evolve the learning of FLCs, especially aiming to the development of fuzzy controllers for mobile robot navigation. The structure has been proven on a Kephera robot. Topics related with the control and automatic navigation of robotic systems especially with learning are approached, based on the Fuzzy Logic theory and evolutionary computing. We can say that our structure corresponds basically to a Classifier System, with appropriate modifications for the objective of generating controllers for mobile robot trajectories. The more stress is made on genetic profile than in the characteristics of the individuals and on the other, the strategy of distribution of the reinforcement is emphasized, fundamental aspects on which the work seeks to contribute.

SFM FOR PLANAR SCENES: A DIRECT AND ROBUST APPROACH

Fadi Dornaika and Angel D. Sappa

Traditionally, the Structure From Motion (SFM) problem has been solved using feature correspondences. This approach requires reliably detected and tracked features between images taken from widespread locations. In this paper, we present a new paradigm to the SFM problem for planar scenes. The novelty of the paradigm lies in the fact that instead of image feature correspondences, only image derivatives are used. We introduce two approaches. The first approach estimates the SFM parameters in two steps. The second approach directly estimates the parameters in one single step. Moreover, for both strategies we introduce the use of robust statistics in order to get robust solutions in presence of outliers. Experiments on both synthetic and real image sequences demonstrated the effectiveness of the developed methods.

PARAMETRIC OPTIMIZATION FOR OPTIMAL SYNTHESIS: of robotic systems’ motions

Taha Chettibi, Moussa Haddad and Samir Hanchi

This paper presents how a problem of optimal trajectory planning, that is an optimal control problem, can be transformed into a parametric optimization problem and in consequence be tackled using efficient deterministic or stochastic parametric optimization techniques. The transformation is done thanks to discretizing some or all continuous system’s variables and forming their time-histories by interpolation. We will discuss three different methods where, in addition to transfer time T, considered optimization parameters are: 1)independent position parameters, 2) control parameters and 3) independent state and control parameters. The simplicity and the efficiency of the first mode allow us to use it to solve the problem of optimal trajectory planning in complex situations, in particular for holonomic and non-holonomic systems.

MULTI-ROBOT SOFTWARE PLATFORM BASED ON ROBOTIC DEVICE SERVER PLAYER

Alejandro Morales, Miguel A. Gutiérrez, Jose A. Vicente, Vidal Moreno and Belén Curto

This article describes a software platform that allows to control multiple robots of any type, through wireless connections and without needing to modify its code to control each particular robot. It is a platform with an architecture in three layers, that uses the robotics device server Player as intermediate layer. The most abstract layer of the architecture is composed by the applications of control elaborated in any language that has socket support. These applications use the interfaces that Player offers to the control of the devices, so that the access to it is transparent. A server application is the most specialized layer that runs on the robot, and it manages the sensors and actuators devices of the robot at Player’s requests. The interesting thing of this platform is that allows to control any robot, without having to develop specific drivers in Player that allow to control their devices. That is to say, it is not necessary to modify the code of the platform to integrate a new robot, simply it is necessary to adapt a model of server application, which accedes to the devices, to the robot that is wanted to integrate. Besides to facilitate the integration of any type of robot on it, the software platform offers the possibility of controlling multiples robots simultaneously through wireless connections (also it admits serial connections). Also, it allows to create valid control programs for any robot, without needing to know its operation and architecture. By these reasons, it constitutes a very valid enviroment to work with multi-robot distributed systems.

CONTROL OF AN ASYMMETRICAL OMNI DIRECTIONAL MOBILE ROBOT

Seyed Mohsen Safavi, Mohammad Ajoodanian and Ahmad movahedian

This paper describes how to control an asymmetric wheeled mobile robot with omnidirectional wheels, considering the sample of a robot with three wheels. When flexible motion capabilities are required this robot must be designed to meet the related requirements, namely fast and agile motions as well as robust navigation. This paper provides an overview for the design of kinematics and dynamics of the robot, as well as motion and velocities equations. In addition to the above a control method to obtain a proper control model is explained. Simulation example is presented to demonstrate the ability of this control method. The implementation and test of the controller on the real robot gives the result compatible with the simulation It was learned that the discontinuities between omnidirectional wheels’ rollers play an important role in decreasing the accuracy of motion.

CALCULATION OF OPTIMAL PATHS IN THE CONFIGURATION SPACE USING ARTIFICIAL POTENTIAL FIELDS AND A* AND D* ALGORITHMS FOR AN ARTICULATED ROBOT. COMPARISON OF TECHNIQUES

Miguel A. Gutiérrez, Alejandro Morales, Vidal Moreno and Belén Curto

In this paper, we use a calculation path technique based on the configuration space in the case of an articulated robot of two degrees of freedom. We propose the use of artificial potential fields to represent the configuration space and the use of techniques of artificial intelligence like A* and D* to search a free collision path into the configuration space. This combination of techniques can be used in static and dynamic environments with more than three dimensions without considering the geometry of the obstacles. The results for this combination of techniques are presented, choosing in each case the best option for each one of the techniques for the combination.

PERSON FOLLOWING BEHAVIOR GENERATED WITH JDE SCHEMA HIERARCHY

Roberto Calvo, José M. Cañas and Lía García-Pérez

One useful capability for service robots is person following. Service robots, like other autonomous robots, are demanded to exhibit a full range of different behaviors. A control architecture is required to integrate, combine and activate all the control mechanisms that generate robot behavior. This paper describes the design and implementation of the person following behavior using JDE. The robot follows a person wearing a coloured shirt, actively searching for her when lost in the image while safely avoiding obstacles. JDE control architecture proposes schemas as basic building blocks for behavior generation and their organization into dynamic hierarchies to scale in complexity. Successful experiments are also described, that validate JDE for this vision-based real-time behavior.

DEVICE SERVER FOR A MINIATURE MOBILE ROBOT

Metin Ozkan and Osman Parlaktuna

This paper describes a device server for a miniature mobile robot. Generally, miniature robots have low-size memory and relatively slow microcontroller to realize complicated tasks. Therefore, a device server for small sized mobile robots is proposed with the intention of increasing their capabilities. The proposed software system runs on the microcontroller of the robot, and serves a collection of sensors and actuators over serial rf transceiver to authorized clients. The system has modularity and multi-tasking capability. The proposed system is implemented on a Z-80 microprocessor-controlled mobile robot. It is shown that proposed system is capable of serving one client and two processes.

VISUALLY SERVOING A GOUGH-STEWART PARALLEL ROBOT ALLOWS FOR REDUCED AND LINEAR KINEMATIC CALIBRATION

Nicolas Andreff and Philippe Martinet

This paper focuses on the benefits of using computer vision to control a Gough-Stewart parallel robot. Namely, it is recalled that observing the legs of such a mechanism with a calibrated camera, thus following the redundant metrology paradigm, simplifies the control law. Then, we prove in this paper that such a control law depends on a reduced set of kinematic parameters (only those attached to the geometry of the robot base) and that these parameters can be obtained by solving a linear system. Moreover, we show that the camera can be calibrated without any experimental effort, simply using images of the robot itself. All this means that setting up the control system consists only in placing the camera in front of the robot.

DESIGN AND DEVELOPMENT OF AUTOMATED SYSTEM FOR LOCALISED ELF MAGNETIC FIELD STIMULATION OF THE HUMAN BRAIN

Dean Cvetkovic and Irena Cosic

The automated system was designed and developed for accurate and fast localisation of extremely low frequency (ELF) electromagnetic field (EMF) exposure to any particular brain region and therefore record for any changes in the EEG activity before and after stimulation. The automated system consisted of a general user interface (GUI) where the users had the ability to precisely control and move an EMF source (coil) via robotic arm to any EEG electrode position or region. The 3-D movements of the robotic arm were controlled via a serial linked motor driver board that controlled two motors. The software was able to initially store the estimated 3-D EEG electrode positions and therefore identify the areas where ELF EMF exposure from the coil could be applied. The testing and final measurements of this system revealed the robotic arm precision of 0.1mm and maximum speed of 0.211 cm/sec (x-axis) and 0.827 cm/sec (y-axis).

RECONFIGURABLE INTERACTIVITY OF PET-TYPE ROBOT REHABILITATION SYSTEM

Toshiyuki Maeda

This paper addresses a pet-type robot rehabilitation system for aged people. The robot offers interactivity, which can communicate autonomously and communicate with others using Internet-connectivity, for being a partner. To avoid being satiated with conversation, we propose reconfigurable interactivity, especially focused conversation contents. In order to watch over aged people through the Net, we have furthermore developed auto-detection alert system for aged people by checking user logs, which is also reconfigurable.

INCREMENTAL LEARNING IN HIERARCHICAL NEURAL NETWORKS FOR OBJECT RECOGNITION

Rebecca Fay, Friedhelm Schwenker and Günther Palm

Robots that perform non-trivial tasks in real-world environments are likely to encounter objects they have not seen before. Thus next to detecting and identifying objects as well as processing language and planning actions the ability to learn new objects is an essential skill for advanced mobile service robots. We have implemented a neurobiologically inspired system on a robot that is able to perform tasks like finding and pointing to certain objects in a complex visual scene according to a spoken or typed command. Furthermore the model has the ability to learn new objects it is shown during run time. This amplifies the adaptability of the approach and thus enables the robot to adjust to new situations. In the following we will concentrate on the object recognition part and on how the incremental learning is implemented there. The intention is to verify whether and how well hierarchical neural networks are suited for extension. The experiments conducted to answer this showed that the proposed incremental learning approach is applicable for hierarchical neural networks and satisfactory classification results can be achieved.

VISION-INERTIAL SYSTEM CALIBRATION FOR TRACKING IN AUGMENTED REALITY

Madjid Maidi, Fakhr-Eddine Ababsa and Malik Mallem

In Augmented Reality (AR) applications, accurate registration of objects is required to project synthetic models at the right location in real images. However, when a vision/inertial hybrid tracker is used, such accuracy depends mostly on the calibration procedure to determine transformations between the sensors frames. This calibration allows to project all data in a single reference frame. In this work, we present a new calibration method for a hybrid tracking system. It consists on rigidly assembling the hybrid tracker to a 6DOF robot in order to simulate the user's head motion while tracking targets in AR environment. Our approach exploits the robot positioning to obtain a high accuracy for the tracker calibration. Experimental results and accuracy analyses are presented and demonstrate our approach effectiveness.

MOMENT BASED FEATURES FOR CONTENT BASED IMAGE RETRIEVAL

Ryszard S. Choras

Current technology allows for the acquisition, transmission, storing, and manipulation of large collections of images. Content based information retrieval is now a widely investigated issue that aims at allowing users of multimedia information systems to retrieve images coherent with a sample image. A way to achieve this goal is the automatic computation of features such as color, texture, shape, and position of objects within images, and the use of the features as query terms. Feature extraction is a crucial part for any such retrieval systems. Current methods of feature extraction suffer from two main problems: firstly many methods do not retain any spatial information and secondly the problem of invariance with respect to standard transformation is still unsolved.\\ In this paper we describe some results of a study on similarity evaluation in image retrieval using shape, texture, color and object orientation and relative position as content features. A simple system is also introduced that computes the feature descriptors and performs queries.\\ Images are retrieval based on similarity of features where features of the query specification are compared with features of the image database to determine which images match similar with the given features. Feature extraction is a crucial part for any of such retrieval systems.

IMAGE BINARISATION USING THE EXTENDED KALMAN FILTER

Alexandra Bartolo, Tracey Cassar, Kenneth P. Camilleri, Simon G. Fabri and Jonathan C. Borg

Form design is frequently carried out by sketches of the designer's mental model of an object. To improve the time it takes from solution concept to production it would therefore be beneficial if paper-based sketches can be automatically interpreted for importation into three-dimensional geometric CAD systems. This however requires image pre-processing before initiating the automated interpretation of the drawing. This paper proposes the application of the Extended Kalman Filter to guide the binarisation process, thus achieving suitable and automatic classification between image foreground and background.

GRASP FEASIBILITY COMPUTATION BASED ON CASCADING FILTERS. APPLICATION TO A THREE FINGERED GRIPPER

Cesar Fernandez, M. Asuncion Vicente, Oscar Reinoso, Luis Paya and Rafael Puerto

A simple, yet effective approach to grasp feasibility analysis is presented. The goal is to reduce the computational complexity of such process, whose complexity makes the detection of all feasible grasps for a certain object unavoidable in most occasions. The approach is based on cascading filters of increasing complexity. First, trivial filters are applied to all the grasp examples, thus rejecting all clearly unfeasible grasps with a small computational effort. Then, more complex filters are applied to a reduced number of grasps and, at the end, the full kinematics and collision detection analysis is only performed with a small subset of the grasps. An example application is presented, where the goal is to detect all the possible 2D grasps of a certain object with an articulated three-fingered hand attached to a scara robot. The vertical axis is decoupled, thus resulting in a highly redundant 7 DOF planar device. Simulation results are presented, where the reduction in computational complexity is evaluated in terms of the number of floating point operations required. Such reduction can be as high as 97\% of the original computation time. An experimental setup has also been developed, with an industrial scara robot and a specifically designed articulated three-fingered gripper. The gripper has pneumatically actuated opening and closing of the fingers and electrically actuated abduction of the articulated fingers. In such experimental setup, the cascading filters approach shows a good behavior. Besides, the proposed system can be easily adapted to different robot arms and hands.

ADAPTIVE VISUAL-FORCE CONTROL IN UNKNOWN WORKSPACES

Jorge Pomares, Fernando Torres and Laura Payá

This paper proposes the definition of a new adaptive system that combines visual and force information. At each moment, the proportion of information used from each sensor is variable depending on the adequacy of each sensor to control the task. The sensorial information obtained is processed to allow the use of both sensors for controlling the robot and avoiding situations in which the control actions are contradictory. Although the visual servoing systems have certain robustness with respect to calibration errors, when the image-based control systems are combined with force control we must accurately know the intrinsic parameters. For this purpose an adaptive approach is proposed which updates the intrinsic parameters during the task.

COMBINING TWO METHODS TO ACCURATELY ESTIMATE DENSE DISPARITY MAPS

Agustín Salgado and Javier Sánchez

The aim of this work is to put together two methods in order to improve the solutions for the problem of 3D geometry reconstruction from a stereoscopic pair of images. We use a method that we have developed in recent works which is based on an energy minimisation technique. This energy yields a partial differential equation (PDE) and is well suited for accurately estimating the disparity maps. One of the problems of this kind of techniques is that it depends strongly on the initial approximation. For this reason we have used a method based on graph--cuts which has demonstrated to obtain good initial guess.

STABILITY ANALYSIS OF A THREE-TIME SCALE SINGULAR PERTURBATION CONTROL FOR A RADIO-CONTROL HELICOPTER ON A PLATFORM

Sergio Esteban, Francisco Gordillo and Javier Aracil

A stability analysis is conducted on the proposed three-time scale singular perturbation control that is applied to a Radio/Control helicopter on a platform to regulate its vertical position. The control law proposed allows to achieve the desired altitude by either selecting a desired collective pitch angle or a desired angular velocity of the blades.

MOTION-EMBEDDED COG JACOBIAN FOR A REAL-TIME HUMANOID MOTION GENERATION

Doik Kim, Youngjin Choi, and ChangHwan Kim

For a legged robot such as a humanoid, balancing its body during a given motion is natural but the most important problem. Recently, a motion given to a humanoid is more and more complicated, and thus the balancing problem becomes much more critical. This paper suggests a real-time motion generation algorithm that guarantees a humanoid to be balanced during the motion. A desired motion of each arm and/or leg is planned by the conventional motion planning method without considering the balancing problem. In order to balance a humanoid, all the given motions are embedded into the COG Jacobian. The COG Jacobian is modified to include the desired motions and, in consequence, dimension of the COG Jacobian is drastically reduced. With the motion-embedded COG Jacobian, balancing and performing a task is completed simultaneously, without changing any other parameters related to the control or planning. Validity and efficiency of the proposed motion-embedded COG Jacobian is simulated in the paper.

MOBILE ROBOT PREDICTIVE TRAJECTORY TRACKING

Martin Seyr and Stefan Jakubek

For a two-wheeled differentially driven mobile robot a trajectory tracking concept is developed. A trajectory is a time-indexed path in the plane, i.e. in the three-dimensional configuration space consisting of position and orientation. Due to the nonholonomic nature of a rolling wheel, the system cannot be stabilized by a continuous time-invariant feedback or by feedback linearization. A novel approach taken in this paper to solve the nonholonomic control problem consists of nonlinear predictive control in conjunction with linear state space control with integration of the control error. Based on a Gauss-Newton algorithm, predicted future position errors are minimized by numerical computation of an optimal sequence of control inputs.

TRACKING-CONTROL INVESTIGATION OF TWO X4-FLYERS

K. M. Zemalache, L. Beji and H. Maaref

This paper presents the study of stabilization with motion planning of two models of four rotors mini-flying robot (helicopter with four rotors rotorcraft). The first is called model inertial without the axes orientation, the second is the model with the axes orientation. Our aim is to obtain algorithms of the control by using the backstepping controllers for the control of this under-actuated system. Our objective is to develop non-linear laws of control able to stabilize the engine in hovering and to generate its trajectory.

KINEMATIC MODELING OF STEWART-GOUGH PLATFORMS

Pedro Cruz, Ricardo Ferreira and João Silva Sequeira

This paper describes a method to solve the direct kinematics of a generic Stewart-Gough manipulators. The method is formulated in terms of a search in the space of rigid body transformations. The underlying idea is that the solutions of the direct kinematics can be obtained by moving the end-effector body according rotations and translations and accounting for the rigidity conditions. The paper presents simulation results for a 6-3 Stewart-Gough robot.

LOCAL PATH PLANNING IN UNKNOWN ENVIRONMENT BY LOCAL 3D ELEVATION MAP CONSTRUCTION

A. Usai and P. Di Giamberardino

The paper deals with the problem of computing a path for an autonomous mobile robot provided by a stereovision camera trough obstacles in an unknown environment with rough ground. The planner makes use of a 3D map reporting the presence and the highness of obstacles together with the shape of the ground and its discontinuities, under the hypothesis of stationarity. The fact that the knowledge of the environment is based only on the images acquired by the cameras, only a local solution can be given unless, or until, the investigation has covered all the working area of the mobile platform. Some experimental or simulation results are used to better clarify each step of the proposed technique.

POSE ESTIMATION OF MOBILE MICROROBOTS IN A SCANNING ELECTRON MICROSCOPE - A cross-correlation based approach using ROI´s

Torsten Sievers and Sergej Fatikow

Mobile microrobots with piezo slip-stick actuation and more than one degree of freedom mostly don't have internal pose sensors. One possibility for fast pose estimation with a high accuracy is the application of video cameras. If accuracy in the micrometer or even in the nanometer range is required, a light microscope respectively a scanning electron microscope (SEM) is needed. In particular the use of a SEM makes high demands on the image processing. High update rates of the pose data enforce a short image acquisition time of the SEM images. Hence the image noise increases, because frame averaging or averaging of the detector signal is time consuming. This paper presents a method to calculate the x,y position and the orientation of a micro gripper in a strongly noised SEM image stream with cross-correlation in real-time. To make real-time pose estimation possible, only a region-of-interest (ROI) is correlated with the target pattern. The SEM is almost predestined to work with ROI's, because the scan area of the electron beam can be chosen arbitrarily. At the beginning of the paper the setup of the used mobile microrobot based nanohandling station will be described briefly.

MULTILAYER PERCEPTRON FUNCTIONAL ADAPTIVE CONTROL FOR TRAJECTORY TRACKING OF WHEELED MOBILE ROBOTS

Marvin K. Bugeja and Simon G. Fabri

Sigmoidal multlayer perceptron neural networks are proposed to effect functional adaptive control for handling the trajectory tracking problem in a nonholonomic wheeled mobile robot. The scheme is developed in discrete time and the multilayer perceptron neural networks are used for the estimation of the robot's nonlinear kinematic functions, which are assumed to be unknown. On-line weight tuning is achieved by employing the extended Kalman filter algorithm based on a specifically formulated multiple-input, multiple-output, stochastic model for the trajectory error dynamics of the mobile base. The estimated functions are then used on a certainty equivalence basis in the control law proposed in (Corradini et al., 2003} for trajectory tracking. The performance of the system is analyzed by simulation.

REAL TIME SIMULATION OF DEFORMABLE OBJECTS WITH FORCE FEEDBACK - Application to surgery simulation

Moulay Brahim El Khalil Ghembaza, Karim Djouani and Yacine Amirat

This paper presents some issues in the simulation of deformable objects with force feedback. It presents an overview of our approach for the conception of a virtual reality medical simulator. We describe a new base finite element method (Extended Tensor-Mass Model) suitable for soft tissue simulation under real time constraints. Our approach allows fast computation of non-linear and viscoelastic mechanical deformations and forces. As far as real-time interactions are concerned, we present our work on collision detection and haptic interaction. Thus, for contact management, a continuous collision detection method based on cubic spline parametric approximation is proposed. Finally, interactive endovascular simulator is described.

PREVISE - A Human-Scale Virtual Environment with Haptic Feedback

Françoix-Xavier Inglese, Philippe Lucidarme, Paul Richard and Jean-Louis Ferrier

This paper presents a human-scale multi-modal virtual environment. User interacts with virtual worlds using a large-scale bimanual haptic interface called the SPIDAR-H. This interface is used to track user’s hands movements and to display various aspects of force feedback associated mainly with contact, weight, and inertia. In order to increase the accuracy of the system, a calibration method is proposed. A large-scale virtual reality application was developed to evaluate the effect of haptic sensation in human performance in tasks involving manual interaction with dynamic virtual objects. The user reaches for and grasps a flying ball. Stereoscopic viewing and auditory feedback are provided to improve user’s immersion.

REMOTE CONTROL OF MOBILE ROBOTS IN LOW BANDWIDTH ENVIRONMENTS

Markus Sauer, Florian Zeiger, Frauke Driewer and Klaus Schilling

Teleoperation of experiments with hardware in tele-education requires information about the working environment and the equipment status as a base. Scenarios with limited bandwidth are of interest for mobile devices as well as for users in areas with a poor telecommunication infrastructure. While camera images provide a realistic view on the remote scene, they need a high bandwidth for quality pictures. In this context an approach to replace transmission of video images is presented. Mixed reality methods are used to visualize the environment, the robot and acquired sensor data. At the example application of tele-learning experiments with mobile robots, data about vehicle position and orientation are essential. This input is to be determined by external tracking systems. The preprocessed sensor information can be sent via internet link even under very low bandwidth conditions. On the students side the robot is visualized in its work space in two- or three-dimensional virtual environments depending on the performance of the available computer. The paper describes the external tracking as well as the remote interface enabling access to the experiments under different conditions and reports about experiences in using this infrastructure.

COMBINING MANUAL HAPTIC PATH PLANNING OF INDUSTRIAL ROBOTS WITH AUTOMATIC PATH SMOOTHING

Heinz Wörn, Björn Hein, Detlef Mages, Berend Denkena, Rene Apitz, Pawel Kowalski and Niels Reimer

Nowadays, industrial robots are preferably programmed offline, i.e. without interference with the real cell and running production processes. Usually a simulation tool is used to manually define individual locations and to check the created raw path for possible collisions. Within this paper an approach is presented, that combines a haptic input device with means of automated path smoothing. The quality of the generated path can be greatly improved by subsequent automatic filtering. Removing redundant locations or modifying intermediate ones increases the smoothness of the path. The semiautomatic programming paradigm with haptic interaction is expected to lead to an improved workflow for robot offline programming.

IMAGE-BASED AND INTRINSIC-FREE VISUAL NAVIGATION OF A MOBILE ROBOT DEFINED AS A GLOBAL VISUAL SERVOING TASK

C. Pérez, N. García-Aracil, J. M. Azorín, J. M. Sabater, L. Navarro and R. Saltarén

The new contribution of this paper is the definition of the visual navigation as a global visual control task which implies continuity problems produced by the changes of visibility of image features during the navigation. A new smooth task function is proposed and a continuous control law is obtained by imposing the exponential decrease of this task function to zero. Finally, the visual servoing techniques used to carry out the navigation are the image-based and the intrinsic-free approaches. Both are independent of calibration errors which is very useful since it is so difficult to get a good calibration in this kind of systems. Also, the second technique allows us to control the camera in spite of the variation of its intrinsic parameters. So, it is possible to modify the zoom of the camera, for instance to get more details, and drive the camera to its reference position at the same time. An exhaustive number of experiments using virtual reality worlds to simulate a typical indoor environment have been carried out.

SOLVING AN INVERSE KINEMATICS PROBLEM FOR A HUMANOID ROBOT’S IMITATION OF HUMAN MOTIONS USING OPTIMIZATION

ChangHwan Kim, Doik Kim and Yonghwan Oh

Interactions of a humanoid with a human are important, when the humanoid is requested to provide people with human-friendly services in unknown or uncertain environment. Such interactions may require more complicated and human-like behaviors from the humanoid. In this work the arm motion of a human is discussed as the early stage of human motion imitation by a humanoid. A motion capture system is used to obtain human-friendly arm motions as references for the humanoid. However the captured motions may not be applied directly to the humanoid, since the differences in geometric or dynamics aspects as length, degrees of freedom, and kinematics and dynamics capabilities exist between the humanoid and the human. To overcome this difficulty a method to adapt captured motions to a humanoid is developed. The geometric difference in the arm length is resolved by scaling the arm length of the humanoid with a constant based on a length ratio. Using the scaled geometry of the humanoid the imitation of actor's arm motion is realized by solving an inverse kinematics problem that is formulated as an optimization problem. The errors between the captured trajectories of actor arms and the approximated trajectories of humanoid arms are minimized. Dynamics capabilities of the joint motors as limits of joint position, velocity and acceleration are also imposed on that optimization problem. Two motions of one hand waving and performing a statement in sign language are imitated by a humanoid in dynamics simulation.

TELE-ROBOTS WITH SHARED AUTONOMY: TELE-PRESENCE FOR HIGH LEVEL OPERABILITY

Thomas Geerinck, Valentin Enescu, Alexandru Salomie, Sid Ahmed Berrabah, Kenny Cauwerts and Hichem Sahli

The aim is to improve the operability of an advanced demonstration platform incorporating reflexive tele-operated control concepts developed on a mobile robot system. The robot is capable of autonomously navigating in semi-structured environments. Reflexive tele-operation mode employs the robot extensive onboard sensor suite to prevent collisions with obstacles when the human operator assumes control and remotely drives the robot to investigate a situation of interest. For the shared autonomy aspect, four levels of autonomy have been implemented: tele-operation, safe mode, shared autonomy and autonomous mode. The operability level is enhanced by improving significantly the situational awareness of the operator by using an inertial tracker in combination with a head mounted display creating a certain feeling of presence. As such, the system permits precision observation and pinpoint data collection without subjecting the user to a possibly hazardous remote environment.

COMPARATIVE PERFORMANCE OF INTELLIGENT IDENTIFICATION AND CONTROL ALGORITHMS FOR A FLEXIBLE BEAM VIBRATION

M. A. Hossain, A. A. Madkour and K. P. Dahal

Petri nets have been widely applied in different aspects of railway modelling and analysis. This paper presents an insight into how coloured Petri nets can be used to model geographical interlocking. We start with a generalisation of coloured Petri nets and follow with an overview of interlocking. In the main body we present a generic unit model and demonstrate how it can be used to represent a simple junction, comprising of three fundamental components; namely track, signal and point units.

A FRAMEWORK FOR TELEPRESENT GAME-PLAY IN LARGE VIRTUAL ENVIRONMENTS

Patrick Rößler, Frederik Beutler and Uwe D. Hanebeck

Immersive games are computer games, in which the user experiences the game world from an avatar’s view. This avatar serves as the user’s alter ego in the game, which these games an excellent testbed for telepresence applications. In this paper we present a framework that provides a novel interface to avatar control in immersive computer games. The user’s motion is tracked and transferred to to the game environment. This motion data is used as control input for the avatar. The game graphics are rendered according to the avatar’s motion and presented to the user on a head-mounted display. As a result, the user immerses into the game environment and identifies with the avatar. However, without further processing of the motion data, the virtual environment would be limited to the size of the user’s real environment, which is not desirable. By using Motion Compression, the framework allows exploring an arbitrarily large virtual environment while the user is actually moving in an environment of limited size. Based on the proposed framework, two game applications were implemented, a modification of a commercially available game and a custom designed game. These two applications proof, that a telepresence system using Motion Compression is a highly intuitive interface to game control.

A GENERIC MODEL FOR ESTIMATING USER INTENTIONS IN HUMAN-ROBOT COOPERATION

Oliver C. Schrempf and Uwe D. Hanebeck

Recognizing user intentions is an important feature for humanoid robots to make implicit and human-like interactions possible. In this paper, we introduce a formal view on user-intentions in human-machine interaction and how they can be estimated by observing user actions. We use Hybrid Dynamic Bayesian Networks to develop a generic model that includes connections between intention, actions and sensor measurements. This model can be used to extend arbitrary human-machine applications by intention recognition.

CONTROL OF DISCRETE LINEAR REPETITIVE PROCESSES WITH VARIABLE PARAMETER UNCERTAINTY

B. Cichy, K. Gałkowski, A. Kummert and E. Rogers

This paper is devoted to solving the problem of stabilising an uncertain discrete linear repetitive process, where the model uncertainty is a result of the variable along the pass uncertainty of the parameters. The analysis is applied to the engineering example of the material rolling process, which can be modelled as a repetitive process (Rogers and Owens, 1992; Gałkowski et al., 2003b). Due to its analytical simplicity and due to computational effectiveness, the LMI based approach to design a robust state controller for 2D systems has been used here.

AN INTELLIGENT VEHICLE APPROACH TO MOBILE VEHICULAR AD HOC NETWORKS - Clustering Optimisation in Dynamic Traffic Networks

James G. Haran, Peng Fan, Peter Nelson and John Dillenburg

The application of Mobile Ad Hoc Network (MANET) technologies to Vehicular Ad Hoc Networks (VANETs) in the service of Intelligent Transportation Systems (ITS) has brought new challenges in maintaining communication clusters of network members for long time durations. Stable clustering methods reduce the overhead of communication relay in MANETs and provide for a more efficient hierarchical network topology. During creation of VANET clusters, each vehicle chooses a head vehicle to follow. Cluster stability in these simulations is measured by the average number of cluster head changes per vehicle during the simulation. In this paper we analyse the effects of six different clustering methods in a simulated highway environment to determine which method provides optimum stability over the simulation timeline.

REMOTE LABORATORY EXPERIMENTS ADDRESSING PATH PLANNING FOR MOBILE ROBOTS

Grzegorz Zyśko, Florian Zeiger, Klaus Schilling and Markus Sauer

This paper describes an educational remote experiment for path planning with mobile robot hardware which is accessible via the internet. The experiment uses a nonholonomic car-like mobile robot with an Ackerman-steering and demonstrates the problems of the inverse kinematics of this kind of mobile robot. It emphasizes the educational aspects, shows how to combine primitive manoeuvres in order to solve the inverse kinematics problem, and gives a detailed description of these manoeuvres.

A SYSTEM FOR TRIDIMENSIONAL IMAGES FROM TWO DIMENSIONAL ONES USING A FOCUSING AND DEFOCUSING VISION SYSTEM

Modesto G. Medina-Meléndrez, David Báez-López, Liliana Díaz-Olavarrieta, J. Rodríguez-Asomoza and L. Guerrero-Ojeda

Machine vision has been made easier by the development of computer systems capable of processing information at high speeds and by inexpensive camera-computer systems. A Camera-Computer system called SIVEDI was developed based in the shape from focusing (SFF) and shape from defocusing (SFD) techniques. The SIVEDI system has as entries the images captured by the camera, the number of steps of the focus mechanism, and the user specifications.

A NEW APPROACH TO AVOID OBSTACLES IN MOBILE ROBOT NAVIGATION: TANGENTIAL ESCAPE

Andre Ferreira, Mario Sarcinelli Filho and Teodiano Freire Bastos Filho

This paper proposes a new strategy for obstacle deviation when a mobile robot is navigating in a semistructured environment. The proposed control architecture is based on a reactive approach, thus demanding low computational effort. It allows the robot to navigate from a starting point to a destination point without colliding to any obstacle in its path. The deviation from an obstacle is performed according to an escape angle calculated so that the new robot orientation is tangent to the obstacle. It is shown that such strategy generates more efcient trajectories, in the sense that the destination point is reached in less time while saving energy and reducing the demand on the robot motors. Another meaningful feature of the proposed strategy is that it also allows to implement the behaviors Wall Following and Corridor Following with no additional computation.

AUTONOMOUS MOBILE ROBOT ASSISTED HERDING

Pinky Thakkar and Leonard Wesley

In this paper, we describe work that begins to address some of the issues related to developing an autonomous mobile robotic capability to assist humans with herding animals. A novel aspect of this work is the development of a capability to convey instructions to the robot via movements of a "toy human." In this work, no other explicit form of communication from the human to the robot is required. Furthermore, the robot is able detect if the human is absent or is unable to herd, and to herd the animal autonomously if required. We developed a herding framework that is based on low stress herding techniques. The robot uses a pan-tilt-zoom camera and a laser ranging sensor to track the human and interpret the human's movements. We conduct two sets of experiments that demonstrate autonomous and co-operative herding behaviour of the robot. We conclude by presenting experimental results that suggest our approach to developing a service robot with assistive herding capabilities holds promise for scaling to more complex and sophisticated tasks.

A STUDY OF CLASSIFICATION TECHNIQUES APPLIED TO CBERS SATELLITE IMAGES

Priscila Andrea da Rocha Severino, Rossana Baptista Queiroz, Arthur Tórgo Gómez and Luiz Paulo Luna de Oliveira

In this paper its presented classification methods for identify forests with araucaria angustifolia, using Artificial Intelligence and Fractal approach. Studies were made to perform experiments in which could be verified the suitable of ANNs for classification of CBERS satellite images. However, it was noticed in that classification a significant error exists. Then, it intends to continuity that study through the incorporation of new techniques of treatment of the images before the submission to Neural Networks training with the use of error minimization techniques. It was observed in the images that some classes has its very defined limits and own characteristics, that it was used one of the technique that it intends to decrease the classification error that happens in the borders, or limit, of transition between a class and another. When applying the detection of borders in those images, it was noticed that those limits have visibly, patterns that could be good as additional information for identification of a class. Therefore, it is supposed that those differences can be quantified by Fractal Dimension calculation, whose definition is going of encounter with the need of establishing patterns for those borders or limits. Fractal Dimension study verifies the adaptation of that technique to determine areas that the Neural Networks and the method Maximum Likelihood doesn't get to distinguish.

PRODUCTION TIME MINIMIZATION STRATEGIES IN A FLEXIBLE MANUFACTURING ENVIRONMENT - A Tabu Search approach

Antonio Gabriel Rodrigues, Arthur Tórgo Gómez and José Vicente Canto do Santos

In this paper it’s proposed a computational model (“Modelo de Seleção de Partes e Escalonamento” – MSPE) to generate a scheduling of parts in a Flexible Manufacturing System environment, considering due dates, production turns and machine tools with magazine constraints. The problems considered are the Part Selection Problem and the Scheduling Problem, the ones which are solved through Cluster Analysis and Tabu Search. The scheduling objectives are the minimization of switching tools total time, minimizing stop instants total time and minimizing the parts tardiness. The optimisation police are defined according to Objective Function’s weights values. Experiments were made to avail the quality of the results obtained with three minimization polices. In these experiments, minimization polices conflicts can be noticed. The definition of Part Families is fundamental to obtain schedules whose reduce the production times.

ROBOT LEARNING BY DEMONSTRATION USING FORWARD MODELS OF SCHEMA-BASED BEHAVIORS

Adam Olenderski, Monica Nicolescu and Sushil Louis

A significant challenge in designing robot systems that learn from a teacher's demonstration is the ability to map the perceived behavior of the trainer to an existing set of primitive behaviors. A main difficulty is that the observed actions may constitute a combination of individual behaviors' outcomes, which would require a decomposition of the observation onto multiple primitive behaviors. This paper presents an approach to robot learning by demonstration that uses a potential-field behavioral representation to learn tasks composed by superposition of behaviors. The method allows a robot to infer essential aspects of the demonstrated tasks, which could not be captured if combinations of behaviors would not have been considered. We validate our approach in a simulated environment with a Pioneer 3DX mobile robot.

A NEW FAMILY OF CONTROLLERS FOR POSITION CONTROL OF ROBOT MANIPULATORS

Fernando Reyes, Jaime Cid, Marco Mendoza and Isela Bonilla

This paper addresses the problem of position control for robot manipulators. A new family of position controllers with gravity compensation for global position of robot manipulators is presented. The previous results on the linear PD controller are extended to the new proposed family. The main contribution of this paper is to prove that closed-loop system composed by full nonlinear robot dynamics and the family controllers is globally asymptotically stable in Lyapunov sense and LaSalle`s invariance principle. Besides the theoretical results, a real-time experimental comparison is also presented to illustrate the performance of the proposed family on a direct-drive pendulum.

CONTINUOUS NAVIGATION OF A MOBILE ROBOT WITH AN APPEARANCE-BASED APPROACH

Luis Payá, M. Asunción Vicente, Laura Navarro, Oscar Reinoso, César Fernández and Arturo Gil

Appearance-based approaches have become a feasible technique applied to robot navigation. They are based on the direct comparison of images without any feature extraction. This approach presents several advantages comparing to model-based methods, such as their application to non-structured environments and the relative simplicity of the control they offer. However, the main drawback of these techniques is the requirement of huge memories and the computational cost they suppose due to the fact that they are based on the continuous comparison of the current images with all those stored in a database. This way, one of the key points for the success of these approaches is the nature and the quantity of information about the environment that must be stored to make possible the navigation. This work presents two alternatives applied the following of pre-recorded routes. The first one consists of storing low-resolution frontal images along the route to follow, acquiring them with a couple of parallel cameras. Several control schemas have been tested to optimize the accuracy in the navigation, such as P, PD and PD with variable parameters. The second alternative consists of reducing the dimension of the data to store, calculating just the most relevant information of each image. We show how it can be done by working in the PCA subspace. This method allows a reduction in the computational cost without necessity of diminishing the resolution of the images, what implies that it could be used in extremely non-structured environments and without reducing the linear speed of the robot.

HARDWARE INDEPENDENT ARCHITECTURE FOR AUTONOMOUS COLABORATIVE AGENTS

Guillermo Glez. de Rivera, Ricardo Ribalda, Kostadin Koroutchev, José Colás and Javier Garrido

We describe a mobile robot test system. A modular structure of the system allows easy inclusion of user hardware modules and communication systems. A client/server approach, XML/RPC based, makes the system easy to program from any operating system & language. The hardware modules are included using a hardware independent protocol as the main feature of the interface between the modules and the system that makes it very flexible and easy to use. The architecture by itself has support for many different communication modalities, as Bluetooth, GSM, Wireless, Ethernet, etc. This architecture was tested including diverse peripheries as for example video camera, microphones, speakers, etc…

A GRAPHICAL INTERFACE BASED ON GRAFCET FOR PROGRAMMING INDUSTRIAL ROBOTS OFF-LINE

Gustavo V. Arnold, Pedro R. Henriques and Jaime C. Fonseca

This paper presents the current development stage of our approach to industrial robot programming, that is the graphical interface for our environment, that is based on the well-known Grafcet. Our approach focus on the modelling of the system, rather than on the robot. So, it will improve the programming and maintenance tasks, allowing the reuse of source code.

A FUZZY CONTROLLER FOR A SPECIAL GLOVE TO A HAND WITH DISABILITIES

Viorel Stoian, Mircea Ivanescu, Elena Stoian and Ionela Iancu

This paper presents a control method for a medical glove with intelligent actuators for a hand with disabilities. The medical glove has got on outer superior face, an intelligent actuator to every finger, which helps it to bend and to grasp different objects and on outer inferior face a force distributed sensor system. The dynamic model of the outer superior face finger is determined and an approximate model is proposed. The two-level hierarchical control is adopted. The upper level coordinator gathers all the necessary information to resolve the distribution force. Then, the lower-level local control problem is treated as an open-chain hyper-redundant structure control problem. The fuzzy rules are established and a fuzzy controller is proposed.

VISUAL SCENE AUGMENTATION FOR ENHANCED HUMAN PERCEPTION

Daniel Hahn, Frederik Beutler and Uwe D. Hanebeck

In this paper we present an assistive system for hearing-impaired people that consists of a wearable microphone array and an Augmented Reality (AR) system. This system helps the user in communication situations, where many speakers or sources of background noise are present. In order to restore the cocktail party effect multiple microphones are used to estimate the position of individual sound sources. In order to allow the user to interact in complex situations with many speakers, a model to estimate the user's attention is developed. This model determines the sounds, which is in the user's focus of attention. It allows the system to discard irrelevant information and enables the user to focus on certain aspects of the surroundings. Based on the user's hearing impairment, the perception of the speaker in the focus of attention can be enhanced, e.g. by amplification or using a speech-to-text conversion. A prototype has been built to test the model and the concept of the system. Currently the prototype is able to locate sound beacons in three-dimensional space, to perform a simple focus estimation, and to present floating captions in the Augmented Reality. The prototype uses an intentionally simple user interface, in order to minimize distractions.

PARTIAL MOTION PLANNING FRAMEWORK FOR REACTIVE PLANNING WITHIN DYNAMIC ENVIRONMENTS

Stéphane PETTI and Thierry FRAICHARD

This paper addresses the problem of motion planning (MP) in dynamic environments. It is first argued that dynamic environments impose a real-time constraint upon MP: it has a limited time only to compute a motion, the time available being a function of the \textit{dynamicity} of the environment. % Now, given the intrinsic complexity of MP, computing a complete motion to the goal within the time available is, in many real-life situations, impossible to achieve. % Partial Motion Planning (PMP) is the answer proposed in this paper to this problem. PMP is a motion planning scheme with an anytime flavor: when the time available is over, PMP returns the best partial motion to the goal computed so far. % Like reactive decision scheme, PMP faces a safety issue: what guarantee is there that the system will never end up in a critical situations yielding an inevitable collision? % The answer proposed in this paper to this safety issue relies upon the concept of Inevitable Collision States (ICS). %~\cite{FrA04} ICS takes into account both the system dynamics and the moving obstacles. By computing ICS-free partial motion, the system safety can be guaranteed. % Application of PMP to the case of a car-like system in a dynamic environment is presented.

AUTONOMOUS MONITORING AND REACTION TO FAILURES IN A TOPOLOGICAL NAVIGATION SYSTEM

V. Egido, R. Barber, M.J.L. Boada and M.A. Salichs

In this paper a system for simultaneous navigation and monitoring with autonomous reaction to failures is going to be presented. This system is part of a complete navigation system called AURON (Autonomous Robot Navigation). The AURON System autonomy is based on the interaction of four main components: the autonomous generation of an environment representation, the planning of a sequence of actions and perceptions which guide the robot from an initial event to a final one, the navigation that converts sequences in real movements and supervises all the process, and the relocalization that allows to place the robot again in the representation. This system has been implemented in a mobile robot control architecture called AD. AD is a two level architecture: deliberative and automatic. The paper is focused in one deliberative skill, the navigation skill.

COORDINATION OF A PROTOTYPED MANIPULATOR BASED ON AN EXPERIMENTAL VISUO-MOTOR MODEL

Renato de Sousa Dâmaso, Mário Sarcinelli Filho, Teodiano Freire Bastos Filho and Tarcisio Passos Ribeiro de Campos

This paper presents a strategy to build an experimental visuo-motor model for a manipulator coupled to a binocular vision system, which discards any previous algebraic model and the calibration of either the manipulator or the vision system. The space spanned by the selected image characteristics is divided in regions, and the estimated visuo-motor model is represented by a matrix of constant elements associated to each one of such regions. Such matrices are obtained in an incremental way, starting from commands of movement and using the measurements of the variations they cause in the image characteristics. Even when partially filled in, the visuo-motor model can be used for coordinating the manipulator in order to get its end-effector closer to an object and to grasp this object. Preliminary results got from the implementation of the proposed strategy in a prototyped manipulator coupled to a binocular vision system are also presented.

MOTION SEGMENTATION IN SEQUENTIAL IMAGES BASED ON THE DIFFERENTIAL OPTICAL FLOW

Flavio de Barros Vidal and Victor Hugo Casanova Alcalde

This work deals with motion detection from image sequences. An algorithm to estimate the optical flow using differential techniques is presented. Noise effects affecting motion detection were taken into account and provisions to minimize it were implemented. The algorithm was developed within the Matlab environment using mex-files to speed up calculations and it was applied to surveillance and urban traffic images. For the considered cases, the results were quite satisfactory.

A MODEL BASED CONTROL OF COMPRESSED NATURAL GAS INJECTION SYSTEMS

Bruno Maione, Paolo Lino and Alessandro Rizzo

Low fuel consumption and low emissions are key issues in modern internal combustion engines design. For this reason, an effective on-line control of the injection process requires the mathematical equations describing the system dynamics. The inherent nonlinearities make the modeling of the fuel-injection system hard to accomplish. Moreover, it is necessary to trade off between accuracy in representing the dynamical behavior of the most significant variables and the need of reducing complexity to simplify the controller design process. In this paper we present a second order lumped parameters model of a Compressed Natural Gas injection system for control system synthesis and analysis. Based on the proposed model, we propose a generalized predictive controller to regulate the injection pressure, which guarantees good performances and robustness to modeling errors. At the same time, the controller structure is simple enough to request a small computational effort.

TERRAIN CLASSIFICATION FOR OUTDOOR AUTONOMOUS ROBOTS USING SINGLE 2D LASER SCANS - Robot perception for dirt road navigation

Morten Rufus Blas, Søren Riisgaard, Ole Ravn, Nils A. Andersen, Mogens Blanke and Jens C. Andersen

Interpreting laser data to allow autonomous robot navigation on paved as well as dirt roads using a fixed angle 2D laser scanner is a daunting task. This paper introduces an algorithm for terrain classification that fuses four distinctly different classifiers: raw height, step size, slope, and roughness. Input is a single 2D laser scan and output is a classification of each laser scan range reading. The range readings are classified as either returning from an obstacle (not traversable) or otherwise it is classified as traversable. Experimental results are shown and discussed from the implementation done with a department developed Medium Mobile Robot and tests conducted in a national park environment.

ACTIVE STEREO VISION-BASED MOBILE ROBOT NAVIGATION FOR PERSON TRACKING

Valentin Enescu, Geert De Cubber, Kenny Cauwerts, Sid Ahmed Berrabah, Hichem Sahli and Marnix Nuttin

In this paper, we propose a mobile robot architecture for person tracking, consisting of an active stereo vision module (ASVM) and a navigation module (NM). The first tracks the person in stereo images and controls the pan/tilt unit to keep the target in the visual field. Its output, i.e. the 3D position of the person, is fed to the NM, which drives the robot towards the target while avoiding obstacles. As a peculiarity of the system, there is no feedback from the NM or the robot motion controller (RMC) to the ASVM. While this imparts flexibility in combining the ASVM with a wide range of robot platforms, it puts considerable strain on the ASVM. Indeed, besides the changes in the target dynamics, it has to cope with the robot motion during obstacle avoidance. These disturbances are accommodated via online estimation of the dynamic system parameters. Robustness against outliers and occlusions is achieved by employing a multi-hypothesis tracking method - the particle filter - based on a color model of the target. Moreover, to deal with illumination changes, the system adaptively updates the color model of the target. The main contributions of this paper lie in (1) devising a stereo, color-based target tracking method using the stereo geometry constraint and (2) integrating it with a robotic agent in a loosely coupled manner.

CENTRALIZED AND DECENTRALIZED OPTIMISATION TECHNIQUES FOR THE FLEXIBLE JOB SHOP SCHEDULING PROBLEM

Meriem Ennigrou and Khaled Ghédira

This paper proposes two Multi-agent approaches based on a tabu search method for solving the flexible Job Shop scheduling problem. The characteristic of the latter problem is that one or several machines can process one operation so that its processing time depends on the machine used. Such a generalization of the classical problem makes it more and more difficult to solve. The objective is to minimize the makespan or the total duration of the schedule. The proposed models are composed of three classes of agents: Job agents and Resource agents, which are responsible for the satisfaction of the constraints under their jurisdiction and an Interface agent. According to the location of the tabu search core, two versions have been proposed. The first one places the optimisation method only on the Interface agent whereas the second associates to each Resource agent its own optimisation process. Different experimentations have also been performed on different benchmarks and results for both versions have been presented.

INFORMATION-BASED INVERSE KINEMATCS MODELING FOR ANIMATION AND ROBOTICS

Mikyung Kim and Mahmoud Tarokh

The paper proposes a novel method for extremely fast inverse kinematics computation suitable for animation of anthropomorphic limbs, and fast moving lightweight manipulators. In the information intensive preprocessing phase, the workspace of the robot is decomposed into small cells, and joint angle vectors (configurations) and end-effector position/ orientation (posture) data sets are generated randomly in each cell using the forward kinematics. Due to the existence of multiple solutions for a desired posture, the generated configurations form a cluster in the joint space. A method involving classification and fuzzy rule base reasoning is described to separate the cluster into appropriate solutions. After the classification, the data belonging to each solution is used to determine the parameters of a simple linear or quadratic model that closely approximates the inverse kinematics within a cell. These parameters are stored in a lookup file. During the online phase, given the desired posture, the index of the appropriate cell is found, the model parameters are retrieved, and the joint angles are computed. The advantages of the proposed method over the existing approaches are discussed in the paper. In particular, the method is complete (provides all solutions), and is extremely fast. Statistical analyses for both an industrial manipulator and an anthropomorphic arm are provided which demonstrate the performance of the proposed method

INTELLIGENT ROBOTIC PERSON FOLLOWING IN UNSTRUCTURED ENVIRONMENTS

Mahmoud Tarokh and John Kuo

The paper describes a scheme based on image identification and fuzzy logic control for following a person by a mobile robot in previously unknown and rough environments. The mobile robot is equipped with a pan-tilt-zoom camera and sonar range sensors. The person detection system uses color and shape of the person to be followed, and provides key characteristics of the person’s image to a fuzzy control scheme. These characteristics are used by fuzzy controllers to determine the actuation signals for the camera pan and tilt, and the robot speed and steering. Experimental results are reported for both indoor locations consisting of tours of labs and hallway, and outdoor environments involving traversal over hills and rough terrain.

KINEMATIC AND SINGULARITY ANALYSIS OF THE HYDRAULIC SHOULDER: A 3-DOF Redundant Parallel Manipulator

H. Sadjadian and H. D. Taghirad

In this paper, kinematic modeling and singularity analysis of a three DOF redundant parallel manipulator has been elaborated in detail. It is known, that on the contrary to series manipulators, the forward kinematic map of parallel manipulators involves highly coupled nonlinear equations, whose closed-form solution derivation is a real challenge. This issue is of great importance noting that the forward kinematics solution is a key element in closed loop position control of parallel manipulators. Using the novel idea of kinematic chains recently developed for parallel manipulators, both inverse and forward kinematics of our parallel manipulator are fully developed, and a closed-form solution for the forward kinematic map of the parallel manipulator is derived. The closed form solution is also obtained in detail for the Jacobian of the mechanism and singularity analysis of the manipulator is performed based on the computed Jacobian.

ADAPTIVE STRATEGY SELECTION FOR MULTI-ROBOT SEARCH BASED ON LOCAL COMMUNICATION AND SENSING

Damien Bright

This paper presents a model for simulating locally adaptive movement strategies for robots involved in collective robotic search tasks in rapidly changing and uncertain environments. The model assumes that the nature of the environment restricts inter-robot communication and uses a stigmergy style local communication mechanism of a type which has been widely applied in collective robots. The basis of the simulation model is a biased random walk where the degree of bias and the step size are both local control variables which change depending on strategies which adapt to the locally sensed environment. The local adaption strategies use an approach based on perceptron style activation functions to control the dominance of weighted estimates of optimal path choice over random choice for robot movement. Experiments aim to test the effectiveness of this approach for optimal collective search in various test domains. A series of initial experiments is presented demontrating the model.

A REFERENCE ARCHITECTURE FOR MANAGING VARIABILITY AMONG TELEOPERATED SERVICE ROBOTS

Francisco Ortiz, Juan Ángel Pastor, Diego Alonso, Fernando Losilla and Esther de Jódar

Teleoperated robots are used to perform hazardous tasks that human operators cannot carry out. The purpose of this paper is to present a new architecture (ACROSET) for the development of these systems that takes into account the current advances in robotic architectures while adopting the component-oriented approach. ACROSET provides a common framework for developing this kind of robotized systems and for integrating intelligent components. The architecture is currently being used, tested and improved in the development of a family of robots, teleoperated cranes and vehicles which perform environmentally friendly cleaning of ship-hull surfaces (the EFTCoR project).

COOPERATIVE MULTI-ROBOT LOCALIZATION: USING COMMUNICATION TO REDUCE LOCALIZATION ERROR

Valguima Odakura and Anna Helena Reali Costa

There has been an increased research interest in localization of a group of mobile robots. In this paper is presented a statistical algorithm for cooperative multi-robot localization. Our approach is based on well-known probabilistic localization approach, the Markov localization. The paper attempts to show that localization error can be decreased and the distance that a robot has to travel in order to find its pose can be diminished when appropriate information is shared among the robots. Robots share their beliefs of poses whenever one robot detects another. We argue that all robots in the group can benefit from a meeting of two robots and we propose a modified detection model with this idea. The technique has been implemented and tested in simulated environments. Experiments illustrate improvements in localization accuracy when compared with a previous multi-robot localization approach.

FILLED Video data based fill level detection of agricultural bulk freight

Fabian Graefe, Walter Schumacher, Raul Queiroz Feitosa and Diogo Menezes Duarte

For automation of a continuous overloading process between two vehicles in motion, two information are essential. On the one hand there is the relative position between the vehicles to be known. On the other hand the loading point within the load space of the transport car has to be determined. Often only a non optimal usage of the transport capacity is obtained without moving the overload swivel. In order to optimize the filling by moving load point the distribution of the freight with in the load space has to be measured during the overload process. In this article the Institut für Regelungstechnik of the Technische Universität in Braunschweig introduces the system FILLED for video data based fill level detection of agricultural bulk freight such as chaffed corn or grass.

COMPARATIVE OF HAPTIC INTERFACES FOR ROBOT-ASSISTED SURGERY

J. M. Azorín, J. M. Sabater, N. García, F. J. Martínez, L. Navarro and R. J. Saltarén

This paper presents a comparative of different non-specific haptic interfaces that could be used for robot-assisted surgery. The purpose of this analysis is to determine which master interface has the best performance for a specific task in which the master-slave scale factor is less than one. Three haptic interfaces have been considered: two commercial masters, one with serial configuration, the PHANToM 1.5 prototype master, and one with spherical setup, the Microsoft Force Feedback 2 Sidewinder; and other one non commercial with a parallel architecture designed in our laboratory, the Magister-P. Two experiments performed to measure the fidelity of the haptic interfaces have been described and the results obtained have been discussed on this paper.

PRIMOS – A NOVEL CONCEPT TO PROGRAM COMPLEX ASSEMBLY PROCESSES

Markus Ehrmann, Jochen Schlick, Marc Seckner and Detlef Zuehlke

Over the past years requirements and size of robot programs have continuously increased. Especially assembly processes increasingly integrate sensors and sensor-based positioning methods to ensure safe processes. Until now programming is realized in manufacturer-dependent text-oriented or graphic-supported simulation systems. If such complex processes have to be realized, both methods result in various disadvantages: Text-oriented programs loose their overview and simulation systems are in need of entire environment models. Due to these reasons, a new concept has been developed in order to improve and simplify the programming of complex sensor based assembly processes. The main objectives of the concept are reducing complexity of robot programs, facilitating clearness for users, supporting diagnostics and handling of trouble during programming. Therefore the technique of visual programming is used and the program is described in an abstract manner by linking graphical symbols. They represent movement of robots and positions of endeffectors. To execute various tasks, so called actions are assigned to the program flow. Further on a concept for handling occurring troubles is integrated. So called exceptions are user-defined and consist of various types of troubles. If an exception is triggered, the program flow will be interrupted and reactions take place. For validation, the concept has been successfully implemented in a tool, named PRIMOS (Programming Robots with an Interference Handling Motion Orientated System). It has been positively evaluated by programming a sensor based assembly process of flanges on optical fibres.

A REACTIVE MOTION PLANNER ARCHITECTURE FOR GENERIC MOBILE ROBOTS BASED ON MULTILAYERED CELLULAR AUTOMATA

Fabio M. Marchese

The aim of this paper is to describe the architecture of a Path Planner for Mobile Robots based on the paradigm of Cellular Automata. The environment representation is distributed, as the robot shape; both and the robot kinematics are parameters for the planner. Hence, it results to be very flexible, handling robots with quite different kinematics (omnidirectional, car-like, asymmetrical, etc.), with generic shapes (even with concavities and holes) and with generic cinematic center positions. Because of these characteristics, it is applicable for the assembly planning in the manufacturing industry, as in the Piano Mover's problems, or in vehicles trajectories generation. It can be applied to flat (Euclidean) Work Space and to natural variable terrains. Considering robots moving with smoothed trajectories, the underlying algorithm is based on a Potential Fields Method, using an anisotropic propagation of potentials on a non-Euclidean manifold. The collision-free trajectories are found following the minimum valley of the potential hypersurface embedded in a 4D space. Thanks to the Multilayered Cellular Automata architecture, it turns out to be very fast, complete and optimal, allowing to react to the wold dynamics (reactive planning), generating new optimal solutions every time the obstacles positions changes.

PRECISE DEAD-RECKONING FOR MOBILE ROBOTS USING MULTIPLE OPTICAL MOUSE SENSORS

Daisuke SEKIMORI and Fumio MIYAZAKI

In this paper, in order to develop an accurate localization for mobile robots, we propose a dead-reckoningsystem based on increments of the robot movements read directly from the floor using optical mouse sensors. The movements of two axes are measurable with an optical mouse sensor. Therefore, in order to calculate a robot's deviation of position and orientation, it is necessary to attach two optical mouse sensors in the robot. However, it is also assumed that a sensor cannot read the movements correctly due to the condition of the floor, the shaking of the robot, etc. To solve this problem, we arrange multiple optical mouse sensors around the robot and compare sensor values. By selecting reliable sensor values, accurate dead-reckoning is realized. Finally, we verify the effectiveness of this algorithm through several experiments with an actual robot.

DEVELOPMENT OF POWER ASSIST ON OMNI-DIRECTIONAL MOBILE WHEELCHAIR CONSIDERING OPERATIONALITY AND COMFORT

Juan Urbano, Kazuhiko Terashima, Takahiro Nishigaki, Takanori Miyoshi and Hideo Kitagawa

In this paper, a power assist system of Omni-directional Mobile Wheelchair(OMW) for helpers aiming at the reduction of incidence by operation of helpers is presented. The OMW presented in this paper, has 3 degrees of freedom, so it is important to consider operationality. And wheelchair's most important performance is passenger's comfort. Therefore, the control system is needed to develop considering both of operationality and comfort. A Power assist controller is proposed and a tuning system of parameters is developed. The effectiveness of the proposed method is demonstrated through experiments.

MASSAGE CONTROL TO ADAPT HUMAN SKIN MUSCLE CONDITION BY USING MULTIFINGERED ROBOT HAND

Kazuhiko Terashima, Taku Kondo, Panya Minyong, Takanori Miyoshi and Hideo Kitagawa

The purpose of this paper is to propose adaptive and flexible expert masssage robot using multi-fingered robot. Towards this goal, the present paper gives a modeling of human skin muscle through robot perception of impedance, and control strategy using impedance control to implement adaptive control system, even if human condition is changed, or massage position is shifted, and person to be massaged is different. The model validity is demonstrated via many experimants by using multi-fingered robot hand and human body. Based on robot perception of human muscle impedance, control strategy for developing an adaptive massage robot and impedance control are proposed by givining the concept of controller switching and sense feedback.

FIELD GEOLOGY WITH A WEARABLE COMPUTER: FIRST RESULTS OF THE CYBORG ASTROBIOLOGIST SYSTEM

Patrick C. McGuire, Javier Gómez-Elvira, José Antonio Rodríguez-Manfredi, Eduardo Sebastián-Martínez, Jens Ormö, Enrique Díaz-Martínez, Markus Oesker, Robert Haschke, Jörg Ontrup and Helge Ritter

We present results from the first geological field tests of the `Cyborg Astrobiologist', which is a wearable computer and video camcorder system that we are using to test and train a computer-vision system towards having some of the autonomous decision-making capabilities of a field-geologist. The Cyborg Astrobiologist platform has thus far been used for testing and development of these algorithms and systems: robotic acquisition of quasi-mosaics of images, real-time image segmentation, and real-time determination of interesting points in the image mosaics. The hardware and software systems function reliably, and the computer-vision algorithms are adequate for the first field tests. In addition to the proof-of-concept aspect of these field tests, the main result of these field tests is the enumeration of those issues that we can improve in the future, including: dealing with structural shadow and microtexture, and also, controlling the camera's zoom lens in an intelligent manner. Nonetheless, despite these and other technical inadequacies, this Cyborg Astrobiologist system, consisting of a camera-equipped wearable-computer and its computer-vision algorithms, has demonstrated its ability of finding genuinely interesting points in real-time in the geological scenery, and then gathering more information about these interest points in an automated manner. We use these capabilities for autonomous guidance towards geological points-of-interest.

HUMAN COGNITIVE SIMULATION FOR EVALUATION OF HUMAN-ROBOT INTERFACE - A trade-off between flexibility in robot control and mental workload

Hiroshi Furukawa

Adaptable automation is a scheme that human operators can modify function allocations among human and machines (or robots) dynamically depending on situations. The critical concept is that operators should be able to delegate tasks to autonomous agents at times of their own choosing, and receive feedback on their performance, just as in successful human-human teams. Playbook is an example of a delegation architecture based on a team’s book of approved plays that provides a “common language” for efficient and effective communication between human operators and agents. Previous studies examined the efficacy of Playbook interface using the Roboflag simulation platform. The results confirmed the benefits, compared to less flexible interfaces which are susceptible to negative effects due to suboptimal automation or unexpected events. This benefit was somewhat reduced, however, when the number of robots was increased. At this higher load, the benefit may have been reduced due to the greater workload demand imposed by full flexibility. This paper described a probabilistic simulation method to estimate behaviors of human operators as a tool for evaluating human-robot interfaces for operation of multiple robots. Through its application to the multiple robots simulation, advantages and costs of different design alternatives has been investigated in terms of cognitive workload indexes of the human operators. Also, the results may suggest the validity of the hypothesis that there is a trade-off between flexibility in operational alternatives and operator’s mental workload.

WHEN SHOULD THE NON LINEAR CAMERA CALIBRATION BE CONSIDERED?

Carlos Ricolfe-Viala and Antonio-José Sánchez-Salmerón

In 3D modelling reconstruction of points, lines, planes or conics are done in the virtual 3D space. Their situations in the 3D virtual scene are defined by the situation of the recognized features in one or several images. Estimation of a parameter vector which models the object is carried out starting with recognized features in the image. Since positions of recognized features in the image are contaminated with noise the solution for the parameter vector is not exact. In order to obtain “the best” solution, optimization algorithms which reduce a residual error are used. They can be classified into linear and non linear ones.The aim of this paper is to determine the quality of estimated parameters if no linear estimation process is utilized. It is shown that in some cases non linear optimization algorithms diverges and worst parameters are computed using non linear methods. In order to obtain experimental results, camera parameters have been estimated under different conditions.

DISTRIBUTED GRADIENT FOR MULTI-ROBOT MOTION PLANNING

Gerasimos Rigatos

Distributed stochastic search is proposed for cooperative behavior in multi-robot systems. Distributed gradient is examined. This method consists of multiple stochastic search algorithms that start from different points in the solutions space and interact to each other while moving towards the goal position. Distributed gradient is shown to be efficient when the motion of the robots towards the goal position is described by a quadratic cost function. The algorithm's performance is evaluated through simulation tests.

EXAMINATION OF BALL TRACKING AND CATCHING TASK USING A MONOCULAR VISION-BASED MOBILE ROBOT

Fumiaki TAKAGI, Fumio MIYAZAKI and Ryosuke MORI

This paper presents an implementation of a ball catching task using a monocular vision-based mobile robot. We have proposed a motion strategy for catching a ball flying in three-dimensional space. This strategy has its roots in the field of experimental psychology but is more powerful and concentrated on a robot. A practical trajectory control law is derived for a non-holonomic mobile robot to track and catch a ball. This control law educes the full potential of the motion strategy: we experimentally demonstrate that a monocular vision-based mobile robot, coping with the problem due to its non-holonomic constraint, successfully catches a ball.

AN IMAGE PROCESSING ALGORITHM - Saving valuable time in a sequence of frames analysis

E. Karvelas, D. Doussis and K. Hrissagis

This paper describes a new algorithm to detect moving objects in a dynamic scene based on statistical analysis of the greyscale variations on a sequence of frames which have been taken in a time period. The main goal of the algorithm is to identify changes (e.g. motion) while coping with variations on environmental changing conditions without being necessary to perform a prior training procedure. In this way, we use a pixel level comparison of subsequent frames in order to deal with temporal stability and fast changes. In addition, this method computes the temporal changes in the video sequence by incorporating statistical results and it is less sensitive to noise. The algorithm’s goal is not to detect motion but rather to filter out similar frames in a sequence of frames, thus making it a valuable tool for those who would like to evaluate and analyze visual information obtained from a captured video frames. Finally, experimental results and a performance measure establishing the confidence of the method are presented.

A GLOVE INTERFACE WITH TACTILE FEELING DISPLAY FOR HUMANOID ROBOTICS AND VIRTUAL REALITY SYSTEMS

Michele Folgheraiter, Giuseppina Gini and Dario L. Vercesi

This paper focuses on the study and the experimentation of a glove interface for robotics and virtual reality applications. The system can acquire the phalanxes position and force of an operator during the execution of a grasp. We show how it is possible to use and integrate this data in order to permit the user to interact with a synthetic world. In particular the system we designed can reproduce tactile and force sensation. Electrodes and actuators are activated according to the information coming from the real world (position and force of the user's finger) and from a physical model that represents the virtual object. We also report some psychophysical experiments we conducted on five subjects, in this case only the electro-tactile stimulator was used in order to generate a touch sensation.

DIRECT GRADIENT-BASED REINFORCEMENT LEARNING FOR ROBOT BEHAVIOR LEARNING

Andres El-Fakdi, Marc Carreras and Pere Ridao

Autonomous Underwater Vehicles (AUV) represent a challenging control problem with complex, noisy, dynamics. Nowadays, not only the continuous scientific advances in underwater robotics but the increasing number of sub sea missions and its complexity ask for an automatization of submarine processes. This paper proposes a high-level control system for solving the action selection problem of an autonomous robot. The system is characterized by the use of Reinforcement Learning Direct Policy Search methods (RLDPS) for learning the internal state/action mapping of some behaviors. We demonstrate its feasibility with simulated experiments using the model of our underwater robot URIS in a target following task.

LOWER LIMB PROSTHESIS: FINAL PROTOTYPE RELEASE AND CONTROL SETTING METHODOLOGIES

Vicentini Federico, Canina Marita and Rovetta Alberto

The current research activity on prostheses project at the Robotics Laboratory (Mechanics Department, Politecnico di Milano) is carried on in cooperation with Centro Protesi INAIL and STMicroelectronics. The team is both innovative and interesting, owing to the fact that it not only involves a range of specialists but also gives rise to interdisciplinary aspects. They are absolutely essential in project dealing with such complex issues. This Mechanics-Leg project, called Hermes, is an original solution in the field of prosthesis. Main aim of this research is the prototyping of a new kind of mechanical lower limb with an electronic control. The device, resorting to innovatory mechanical and electronic solutions, allows the controller to modify the type of step, passing from a slow to a fast walk, in an easy and intuitive way, taking care of patient’s requirements. The Hermes M-Leg cost is comparable to the actual commercial non electronic controlled artificial knees. The distinguishing features of M-Leg Hermes project are an higher awareness in innovative aspects related to medical/biological/engineering research. Then, a pervasive use of cutting-edge technology (electronics, IT, material-related technologies, etc.). The controller architecture is built upon a low memory processing features. The hard analysis and test activity help to model the algorithm for step control. The adaptive behaviour is mostly due to an effective experience in testing and software tuning in cooperation with patients and clinical staff.

A SMOOTHING STRATEGY FOR PRM PATHS: Application to 6-axes MOTOMAN manipulator

R. Guernane and M. Belhocine

This paper describes the use of the probabilistic motion planning technique SBL “Single-Query Bidirectional Probabilistic Algorithm with Lazy Collision Checking” or in motion planning for robot manipulators. We present a novel strategy to remedy the PRM “Probabilistic Roadmap” paths which are both excessively long and discontinuous in velocity and acceleration. The optimization of the path will be done first through Coarse optimal lazy A* optimization and then through a Fine cutting-triangles-edge one, the edges discontinuities are smoothed with cubic polynomials taking the robot’s specific Dynamic and Cinematic constraints. The whole strategy is applied to the 6 axes robot Manipulator MOTOMAN SV3X.

AFFORDABLE DEEP OCEAN EXPLORATION WITH A HOVERING AUTONOMOUS UNDERWATER VEHICLE - Odyssey IV: a 6000 meter rated, cruising and hovering AUV

V. Polidoro, S. Desset, C. Chryssostomidis, F. Hover, J. Morash and R. Damus

The Autonomous Underwater Vehicle Laboratory (AUV Lab) at The Massachusetts Institute of Technology (MIT) is currently building and testing a new, general purpose and inexpensive 6000 meter capable Hovering Autonomous Underwater Vehicle (HAUV), the ‘ODYSSEY IV class’. The vehicle is intended for rapid deployments, potentially with minimal navigation, thus supporting episodic dives for exploratory missions. For that, the vehicle is capable of fast dive times, short survey on bottom and simple navigation. This vehicle has both high speed cruising and zero speed hovering capabilities, enabling it to perform both broad area search missions and high resolution inspection missions with the same platform.

APPLICABILITY OF FACIAL EMG IN HCI AND VOICELESS COMMUNICATION

Sanjay Kumar, Dinesh Kant Kumar and Melaku Alemu

This paper discusses the speech related information in the facial EMG for applications such as human computer interface. The primary objective of this work is to investigate the use of facial EMG as a voiceless communication medium or to drive computer based equipment by people who are unable to speak. Subjects were asked to pronounce the five English vowels with no acoustic output (voiceless). Three independent EMG signals were acquired from three facial muscles as ‘voiceless’ EMG activations. In order to classify and recognize each vowels based on EMG, RMS of the recorded signals were estimated and used as parametric/feature inputs to a neural network.

SUS A NEW GENERATION THINKING ROBOTS - The Visual Intelligence Tests

Zbigniew Les and Magdalena Les

In this paper understanding abilities of the shape understanding system (SUS) are tested based on the methods used in the intelligence tests. These tests are formulated as tasks given to the system and performance is compared with the human performance of these tasks. The tests were based on the progressive matrices test which requires the good visual problem solving abilities of the human subject. SUS solves these tests by transforming the visual form into the string form. The proposed string form makes it possible to perform complex visual reasoning. The experiment proved that the proposed method, which is part of the SUS visual understanding abilities, can solve the test that is very difficult for human subject.

A GENERAL SOLUTION TO THE OUTPUT-ZEROING PROBLEM FOR DISCRETE-TIME MIMO LTI SYSTEMS - Signal Processing, Systems Modelling and Control

Jerzy Tokarzewski, Lech Sokalski and Andrzej Muszyński

The problem of zeroing the output in an arbitrary linear discrete-time system S(A,B,C,D) with a nonvanishing transfer-function matrix is discussed and necessary conditions for output-zeroing inputs are formulated. All possible real-valued inputs and real initial conditions which produce the identically zero system response are characterized. Strictly proper and proper systems are discussed separately.

A GRAPHICAL REVIEW OF NOISE-INSTABILITY CHARACTERIZATION IN ELECTRONIC SYSTEMS

Juan José González de la Rosa, Isidro Lloret Galiana, Carlos García Puntonet and Víctor Pallarés López

A thorough study of the noise processes characterization is made with simulated data by means of our non-classical estimators. Individual and hybrid noise sequences, previously generated by seed functions, have been used to obtain a set of characterization graphs identifying the noise type by mean of the enveloping curve. It is also shown the case of a hidden noise. An real test situation is presented which involves a traceable characterization via a GPS receiver.

SUNSPOT SERIES PREDICTION USING ADAPTIVE IDENTIFICATION

Juan A. Gómez Pulido, Miguel A. Vega Rodríguez, José Mª Granado Criado and Juan M. Sánchez Pérez

In this paper a parallel and adaptive methodology for optimizing the time series prediction using System Identification is shown. In order to validate this methodology, a set of time series based on the sun activity measured during the 20th century have been used. The prediction precision for short and long term improves with this technique when it is compared with the found results using System Identification with classical values for the main parameters.

MODELING SYSTEM VARIATION

Ken Krechmer

This paper proposes that the mathematical relationship between an entropy distribution and its limit offers some new insight into system performance. This relationship can be used to quantify variation among the entities of a system, caused by tolerance, options, specification or implementation errors, independent of noise, impact communications system performance. Means to address these variations are offered.

APPLYING SIGNAL PROCESSING TECHNIQUES TO WATER LEVEL ANOMALY DETECTION

Carl Steidley, Richard Rush, David Thomas, Phillipe Tissot, Alex Sadovski and Ray Bachnak

Abstract: The Texas Coastal Ocean Observation Network (TCOON) consists of more than 50 data gathering stations located along the Texas Gulf coast from the Louisiana to Mexico borders. Data sampled at these stations include: precise water levels, wind speed and direction, atmospheric and water temperatures, barometric pressure, and water currents. The measurements collected at these stations are often used in legal proceedings such as littoral boundary determinations; therefore data are collected according to National Ocean Service standards. Some stations of TCOON collect parameters such as turbidity, salinity, and other water quality parameters. All data are transmitted back to Texas A&M University Corpus Christi (A&M-CC) at multiples of six minutes via line-of-sight packet radio, cellular phone, or GOES satellite, where they are then processed and stored in a real-time, web-enabled database. TCOON has been in operation since 1988. This paper describes a software project based upon signal processing techniques to be utilized with the TCOON meteorological database to detect spikes in water level. Water level readings are frequently victim to abnormal water levels caused by ship wakes, affected equipment scrambled by thunder, or corrupted by transmission errors. Since these water levels are the bases for a number of research calculations, such as, oil-spill response, navigation safety, environmental research, and recreation, it is essential to be able to make these water level data as correct and spike free as possible.

INFORMATION SYSTEMS SUPPORT ON MOBILE DEVICE PLATFORM - Java SCADA Client/Server model and .NET localization enhancement

Ondřej Krejcar and Jindřich Černohorský

The paper deals with programming possibilities of mobile devices. It discusses the relationships with control systems and problems with solutions of possible situations arising from their design or their operation. It is focused mainly on Java language and use of created SCADA based application on wide scale of mobile devices without any changes of source code. The proliferation of mobile computing devices and local-area wireless networks has fostered a growing interest in location-aware systems and services. Another area of interest is in model of radio-frequency (RF) based system enhancement for locating and tracking users of our information system inside buildings. User location is used to data pre-buffering and pushing information from server to user’s PDA.

HIGH TEMPERATURE DENSITY MEASUREMENT CELL WITH A PCMCIA-INTERFACE

Bernd Eichberger and Anton Scheibelmasser

One of the most precise and reliable measurement methods for density measurement of liquids and gases depends on the principles of a mechanical oscillator. With this method the density is determined by measuring the natural frequency of the oscillator. Measurement devices using this method can be categorized in two groups. The first type incorporates the mechanical oscillator in the housing of the device and is mainly used in laboratories. The second type of measurement devices could be defined as evaluation units, because the sensor e.g. mechanical oscillator is external and only connected by means of electrical connections. These types are used in the field of process data control or data acquisition. The reason for separating the sensor from the evaluation unit lies in the fact that such external cells are used on remote locations in the process or the sensor is exposed to extreme physical conditions (e.g. high pressure, high temperatures). The first part of this article gives an overview about the functionality of such a high temperature measurement cell. The second part of this paper is intended to introduce a sophisticated PCMCIA interface which acts as an interface between the external density measurement cell and several hosts like PCs, PDAs or modern density measurement devices.

REAL TIME WEB AVAILABILITY OF STATISTICAL MODELS FOR WATER LEVELS ALONG THE TEXAS COASTLINE

Alex Sadovski, Carl Steidley, Philippe Tissot and G. Beate Zimmer

Water level forecasts are essential to the success of trade and industry in the Gulf of Mexico, but present forecasting methodologies do not provide accurate predictions for the Gulf Coast region. Tide charts produced by harmonic analysis are the existing standard, but these charts only show the effect of astronomical forces acting upon the water. While this proves to be an accurate predictor for most of the Atlantic and Pacific Coasts, water level changes along the Texas Coast are strongly affected by meteorological factors and thus require a modified prediction model, rather than harmonic analysis alone. A web-based tool was created that combines harmonic analysis with multivariate statistical modeling to predict water levels along the Texas Gulf Coast. The result is a substantial improvement on the current model with forecasts available via the World Wide Web.

EVALUATION OF TRACTION POWER CONSUMPTION CONTROL SYSTEM IN THE CZECH REPUBLIC - And its Basic Components

Jindrich Sadil, Zuzana Belinova, Vaclav Vodrazka, Jindrich Krasa, Jakub Rajnoch and Petr Bouchner

Our work should help to the higher effectiveness of traction power consumption. The traction power for railway comes from the regional distribution companies of the electric energy in the Czech Republic. The aim is to make cost connected with given criteria of power consumption lower. For example in the Czech Republic it is a cost connected with breach of conditions given by the “Price Decision of the Czech Energetic Regulation Office (ERU) no. 10/2004”. But the scale of our work should be much larger, because each state has its specific conditions of power consumption and the work can solve universal case. It could be done by means of the system, which would use the new knowledge of informatics, telematics and system engineering. This paper discusses the state at this time at this field nowadays and possible application in the territory of ex-district North Moravia and Silesia, which is a part of the Czech Republic. This area is an integrated part of the country, which electricity is separately accounted for.

PERFORMANCE ANALYSIS OF TIMED EVENT GRAPHS WITH MULTIPLIERS USING (Min, +) ALGEBRA

Samir Hamaci, Jean-Louis Boimond and Sébastien Lahaye

We are interested in the performance evaluation of timed event graphs with multipliers. The dynamical equation modelling such graphs are nonlinear in (min,+) algebra. This nonlinearity is due to multipliers and prevents from applying usual performance analysis results. As an alternative, we propose a linearization method in (min,+) algebra of timed event graphs with multipliers. From the obtained linear model, we deduce the cycle time of these graphs. Lower and upper linear approximated models are proposed when linearization condition is not satisfied.

IDENTIFICATION AND PREDICTION OF MULTIPLE SHORT RECORDS BY DYNAMIC BAYESIAN MIXTURES

Pavel Ettler and Miroslav Kárný

A short data record is not suitable for proper identification of system model which is necessary for reliable data prediction. The idea consists in utilization of multiple similar short data records for identification of a dynamic Bayesian mixture. The mixture is used for prediction according to one of three methods described. Simulated and real data examples illustrate the methods.

ON THE LINEAR LEAST-SQUARE PREDICTION PROBLEM

R. M. Fernández-Alcalá, J. Navarro-Moreno, J. C. Ruiz-Molina and M. D. Estudillo

An efficient algorithm is derived for the recursive computation of the filtering and all types of linear least-square prediction estimates (fixed-point, fixed-interval, and fixed-lead predictors) of a nonstationary signal vector. It is assumed that the signal is observed in the presence of an additive white noise which can be correlated with the signal. The methodology employed only requires that the covariance functions involved are factorizable kernels and then it is applicable without the assumption that the signal verifies a state-space model.

A DESIGN METHOD OF TWO-DIMENSIONAL LINEAR PHASE FIR FILTERS USING FRITZ JOHN’S THEOREM

Yasunori Sugita and Naoyuki Aikawa

This paper presents a design method of 2-dimensional (2-D) FIR filters by successive projection (SP) method using multiple extreme frequency points based on Fritz John's theorem. The proposed method enables an update of coefficients using multiple extreme frequency points by Fritz John's theorem. Moreover, we also present two methods as how to choose the extreme frequency point for the update coefficients. As a result, the solution converges less iteration number and computing time than the previous method.

IDENTIFICATION OF A CAR-LIKE VEHICLE via MODULATING FUNCTIONS

Davide Corsanini and Fabrizio Tocchini

This paper describes an interesting application of the modulating functions technique to model identification of a car-like vehicle that has to face various types of soil. Several models have been obtained in different operating conditions. The construction of a `mean model' will make possible the design of a robust control for unmanned guidance purposes. An alternative control strategy based on adaptive methods is also suggested by means of an online implementation of the technique.

PARAMETER ESTIMATION OF MOVING AVERAGE PROCESSES USING CUMULANTS AND NONLINEAR OPTIMIZATION ALGORITHMS

M. Boulouird, M. M. Hassani and G. Favier

In this paper nonlinear optimization algorithms, namely the Gradient descent and the Gauss-Newton algorithms, are proposed for blind identification of MA models. A relationship between third and fourth order cumulants of the noisy system output and the MA parameters is exploited to build a set of nonlinear equations that is solved by means of the two nonlinear optimization algorithms above cited. Simulation results are presented to compare the performance of the proposed algorithms.

DETECTABILITY AND DIAGNOSABILITY OF DISCRETE EVENT SYSTEMS - Application on manufacturing systems

Moamar Sayed Mouchaweh, Alexandre Philippot and Véronique Carré-Ménétrier

The diagnosis is defined as the process of detecting an abnormality in the system behavior and isolating the causes or the sources of this abnormality. Not all the systems are diagnosable. Thus, before Appling a method to diagnose a system, we need to know if this system is diagnosable according to the set of failures required to be detected and isolated. This paper presents an algorithm to determine if a system is detectable or not, i.e., if we can know, at each instant, whether the system works under a normal or abnormal functioning state. In the case that the system is detectable, this algorithm determines if this system is diagnosable. This algorithm combines event and state based approaches in order to maximise the diagnosability power with a minimum number of sensors. In addition the time is integrated and modelled with fuzzy intervals to enhance this diagnosabilty power and to take into account the imprecision of events occurrences instants. An example of manufacturing system is used to illustrate the functioning of this algorithm.

3D AUTOMATIC LOCATION DETECTION BASED ON SOUND LOCALIZATION

Darun Kesrarat and Paitoon Porntrakoon

Video conference systems have been widely used. A fix video camera shoots a scene is lacking in changes. There is a method that the computer-controlled camera shoots and finds the sound source. Microphone arrays and distributed microphone arrays are used to localize the sound source based on time delay of arrival (TDOA). In order to minimize the error rate of TDOA, a set of 4 microphone arrays can be used to determine the location of sound in 3D space. TDOA cannot determine the distance of the sound source if the start time of the sound is unknown. A method to determine the distance of the sound source is using a distributed moving-microphone array. In this paper, we propose a model of a set of 4 moving-micorphone array based on TDOA that can determine the angle direction and distance of the sound source toward the video camera at the center of the model in 3D space.

DECENTRALIZED SLIDING MODE CONTROL TECHNIQUE BASED POWER SYSTEM STABILIZER (PSS) FOR MULTIMACHINE POWER SYSTEM

Vitthal Bandal, B. Bandyopadhyay and A. M. Kulkarni

Power System Stabilizers (PSSs) are added to excitation system to enhance the damping of low frequency oscillations. In this paper, the design of PSS for multimachine power system (MMPS) using output feedback sliding mode control is proposed. The non-linear model of a multimachine power system is linearized about an operating point and the linearized model of the plant is obtained. The output feedback sliding mode controller is designed and is applied to non-linear plant model of the multimachine power system at that operating (equilibrium) point. This method does not require the complete states of the system for feedback and is easily implementable.

MUSICAL INSTRUMENT ESTIMATION FOR POLYPHONY USING AUTOCORRELATION FUNCTIONS

Yoshiaki Tadokoro and Koji Tanishita

This paper proposes a new musical instrument estimation of polyphony using autocorrelation functions. We notice that each musical instrument has each autocorrelation function. Polyphony can be separated into each monophony using comb filters ( ). We can obtain the autocorrelation functions for the outputs of comb filters from the autocorrelation functions of the monophony. By the pattern patching between the autocorrelation functions for the output signals of the comb filters and ones calculated from monophony of each instrument, we can estimate the musical instruments for polyphony.

A NOVEL ENTROPY METHOD FOR CLASSIFICATION OF BIOSIGNALS

Andrea Casanova, Valentina Savona and Sergio Vitulano

The paper introduces entropy as a measure for 1D signals. We propose as entropy measure the relationship between the crest of the signal (i.e. its portion contained between the absolute minimum and maximum) and the energy of the signal. A linear transformation of 2D signals into 1D signals is also illustrated. The experimental results are compared to several fuzzy entropy measures and other well-known methods in literature. Experiments have been carried out on medical images from a large mammograms database; this choice is due to the high-degree of difficulty of this kind of images and the strong interest in the scientific community on medical images. The capability of the methods was tested in order to discriminate between benignant and malignant microcalcifications.

REMOTE MONITORING DISTRIBUTED SYSTEMS - The new generation of control systems

Víctor Ruiz Valera, Mario de la Cruz Ortiz, Rafael Herradón Díez and Florentino Jiménez Muñoz

Nowadays, it is increasingly necessary and interesting to measure and control the levels of some parameters. This research and project work have been developed with the aim of creating a data acquisition and monitoring distributed system that allows the users to monitor and control easily, powerfully and flexibly any parameter interesting enough to be studied for later monitoring in real time. At the same time, it is also intended to make the system accessible to the general public and citizens, by using the most implanted and widespread network: Internet and the TCP/IP networks. As a result we present an application offering a “measurement transport layer” providing several services that will work with any kind of parameter.

MODELING AND CONTROLLER DESIGN OF A MAGNETIC LEVITATION SYSTEM WITH FIVE DEGREES OF FREEDOM

E. Alvarez-Sanchez, Ja. Alvarez-Gallegos and R. Castro-Linares

In this paper, the nonlinear mathematical model with five DOFs (degrees-of-freedom) of a magnetic levitation system is developed and analyzed. Then a second order sliding mode controller is proposed to regulate the levitation to a desired position, stabilize the other 4 DOFs in the nonlinear system and compensate the unknown increments on the load. Simulation results are presented to show the effectiveness of the proposed controller.

FREEZING ALARM SYSTEM BASED ON TIME SERIES ANALISYS

Carmen Morató, M. T. Castellanos, A. M.Tarquis and Enriqueta G. Mouton

The aim of this work is to design an alarm system that allows protecting and preventing crop-freezing damages taking decisions with enough time to react. A first step was to obtain a temperature forecast mode. In this line an hourly temperature series was analyzed with Box-Jenkins methodology ( ARIMA models). An alarm system is designed based on these forecast, at each 12 hours, in the air temperatures obtained each hour at real time and in the average errors between real and forecast each hour and each 12 hours. This system generates an index alarm that is related with the risk intensity that over a certain value will activate several sensors. This system is applicable to any area adjusting conveniently the parameters and the ARIMA model.

MULTI-BAND GPS SIGNAL TRACKING IN A HIGH DYNAMIC MANEUVERING SITUATION

Stanislas Boutoille, Serge Reboul and Mohammed Benjelloun

In a GPS receiver, the goal of the signal tracking is to synchronize local generated code and carrier with the received signal. After a step of acquisition, the receiver tracks the shifting of the local code provoked by the movements of the receiver and satellites. In the future evolution of the GPS, the system will have several carrier frequencies, then it will be possible to have several tracking systems working simultaneously for a same satellite. We present in this article a detection method for the tracking of the future multi-band GPS signal. This method is applied to the localization of a vehicle which makes high dynamic maneuver. We define a MAP detection criterion to fuse the shifts discriminator detection achieved on multi carrier frequencies. This criterion is defined in the case when shifts are not necessary simultaneous and when there is a lack of information on one frequency provoked by the unlocking of the code tracking on one of the carrier. Indeed, there is a difference between the instants of shifts on the different carrier frequencies. This difference is due to the effect of ionospheric propagation. The experimentations achieved on synthetic GPS signals show the advantages of the method compared to the classical algorithm.

MULTI-OBJECTIVE PREDICTIVE CONTROL: APPLICATION FOR AN UNCERTAIN PROCESS

Anes Bedoui, Faouzi Bouani and Mekki Ksouri

This paper deals with the application of the Multi Objective Generalized Predictive Control (MOGPC) to level control in a laboratory process. The major characteristic of the considered plant is that the manual draining vane can take many positions causing changes in plant dynamics and strong disturbances in the process. The controller is based on a set of Controlled Auto Regressive Integrated Moving Average (CARIMA) model. The Recursive Least Squares (RLS) algorithm is used to estimate each model parameters. The control law is obtained by minimizing a multi objective optimization problem. The weighting sum approach is considered to formulate the control problem as a single criterion optimisation one. The real time control system implementation confirms the opportunity of using the MOGPC scheme to an uncertainty system.

A NEW HIERARCHICAL CONTROL SCHEME FOR A CLASS OF CYCLICALLY REPEATED DISCRETE-EVENT SYSTEMS

Danjing Li, Eckart Mayer and Jörg Raisch

We extend the hierarchical control method in \cite{Li04} to a more generic setting which involves cyclically repeated processes. A hierarchical architecture is presented to facilitate control synthesis. Specifically, a conservative max-plus model for cyclically repeated processes is introduced on the upper level which provides an optimal online plan list. An enhanced min-plus algebra based scheme on the lower level not only handles unexpected events but, more importantly, addresses cooperation issues between sub-plants and different cycles. A rail traffic example is given to demonstrate the effectiveness of the proposed approach.

ACOUSTIC NOISE SUPPRESSION: COMPROMISES IN IDENTIFICATION AND CONTROL

Ricardo S. Sánchez Peña, Miquel A. Cugueró, Albert Masip, Joseba Quevedo and Vicenç Puig

A control-oriented robust identification procedure which takes into account both, parametric and non-parametric models, is applied to the primary and secondary circuits of an acoustic tube. These models are used to design an H-infinity optimal controller for noise suppression. The analysis of the closed loop system is performed via the structured singular value (mu). The compromises in the identification and control stages in terms of performance vs. controller order are explicitly pointed out.

GENERAL ENGINEERING DATA MODEL IN SPECIAL PURPOSE MACHINE ENGINEERING

Zhiliang Qi, Christian Schäfer and Peter Klemm

A main problem in todays engineering of special purpose machines is the reuse and the consistency of engineering data across the whole development process. This paper presents an approach how to manage and share data in the entire development process in the field of special purpose machine engineering. The first part gives a short overview of the current problems by using software tools in engineering and the engineering requirements. In the second part the current data models used in special purpose machine engineering are analyzed. The third part provides the General Engineering Data Model (GED) as a new concept to share and reuse the engineering data from each engineering phase and to improve the development activity. At the end this paper gives also an evaluation on benefits and contribution of this GED.

A HYBRID CONTROLLER FOR A NONHOLONOMIC CAR-LIKE ROBOT

Martin v. Mohrenschildt

This paper presents the usage of hybrid systems to develop an adaptive recent horizon controller for a nonholonomic car-like robot. The system is modeled by a non-deterministic hybrid system in which transitions represent the discrete control actions, mode invariants the constraints and the transition relation encodes sequencing requirements. The control algorithm examines at runtime the possible traces into the future by determining at which time point to switch to which mode. Based on these predictions the next control move is performed. We demonstrate our approach by controlling a car-like robot though a maze without any pre-runtime path planing.

ANALYSIS AND SYNTHESIS OF DIGITAL STRUCTURE BY MATRIX METHOD

B. Psenicka, R. Bustamante Bello and M. A. Rodriguez

This paper presents a general matrix algorithm for analysis and synthesis of digital filters. A useful method for computing the state-space matrix of a general digital network and a new technique for the design of digital filters are shown by means of examples. The method proposed in this paper allows the analysis of the digital filters and the construction of new equivalent structures of the canonic and the non canonic digital filter forms. Equivalent filters with different structures can be found according to various matrix expansions. The procedure proposed in this paper is more efficient and economic than traditional methods because it permits to construct circuits with a minimum of shifting operations.

PIECEWISE AFFINE SYSTEMS CONTROLLABILITY AND HYBRID OPTIMAL CONTROL

Aude Rondepierre

We consider a particular class of hybrid systems, defined by a piecewise affine dynamic over non-overlapping simplicial regions of the state space. We want to control their behaviors so that it reaches a target by minimizing a given cost. We provide a new numerical algorithm under-approximating the controllable domain under the given hybrid dynamic. Given an optimal sequence of states of the hybrid automaton, we are then able to traverse the automaton till the target, locally insuring optimality.

DYNAMIC HYSTERESIS MODEL DERIVATED FROM LuGre MODEL

Sinuhe Benitez, Leonardo Acho and Ricardo Guerra

This paper presents a dynamic hysteresis model; which is a modification of the well known LuGre model. This model has been based on a dynamic modification, which could be seen as a forward and backward displacement in the stationary state solution of the dynamic LuGre model. The LuGre friction model is based on the average deflection of the bristles; implicitly, it is based on the relationship between stress and strain of the bristles under deformation. From the friction model point of view, this dynamic hysteresis model can capture the deformation behavior between stress and strain beyond the elasticity zone for the material (the bristles), a zone where the relationship between stress and strain is not longer linear. So, our model can capture the friction phenomena of the original LuGre model and presents a new behavior in the pre-sliding regime. Simulation results are presented to support our contribution.

MODELING AND ANALYSIS OF REDUNDANCY IN REMOTE MONITORING AND CONTROL SYSTEMS VIA PETRI NETS

Cheng Guo and Zheng Qin

This paper presents techniques that enable the modeling and analysis of redundancy in remote monitoring and control systems. Hardware redundancy and software redundancy are both implemented in the proposed system organically. To model and analyze the redundancy, a formalism to derive Petri net (PN) model from state transition diagram is constructed and the monitor compatible condition is considered. Software redundancy mechanism in manager side is also modeled based on PN. This approach makes redundancy modeling and analysis possible in terms of well-developed concepts and methods in PN theory. By the analysis of the PN model, designers can obtain reliable and effective measures that would compare different schema in the early phases of design, and select the best one. Our approach is illustrated and validated on STPNPlay by an example.

ANN-BASED MULTIPLE DIMENSION PREDICTOR FOR SHIP ROUTE PREDICTION

Tianhao Tang and Tianzhen Wang

This paper presents a new multiple dimension predictive model based on the diagonal recurrent neural networks (PDRNN) with a combined learning algorithm. This method can be used to predict not only values, but also some points in the multi-dimension space. And also its applications in data mining will be discussed in the paper. Some analysis results show the significant improvement to ship route prediction using the PDRNN model in database of geographic information system (GIS).

METHOD FOR ALARM PREDICTION

Luis Pastor Sanchez Fernandez, Lazaro Gorostiaga Canepa and Oleksiy Pogrebnyak

The goal of this paper is to show a predictive supervisory method for the trending of variables of technological processes and devices. The data obtained in real time for each variable are used to estimate the parameters of a mathematical model. This model is continuous and of first-order or second-order (critically damped, overdamped or underdamped), all of which show time delay. An optimization algorithm is used for estimating the parameters. Before performing the estimation, the most appropriate model is determined by means of a backpropagation neural network (NN) previously trained. Virtual Instrumentation was used for the method programming.

IDENTIFICATION OF STRUCTURE IN NONDETERMINISTIC CYCLIC SOCIAL CONVENTIONS

Hürevren Kılıç

A polynomial-time algorithm for the identification of interaction and memory structures in discrete valued, nondeterministic, cyclic social behavior data is developed. The output of the probabilistic search algorithm is the strategy update function for each individual automaton agent in given population. For our modeling purpose, we used automata networks model and added “block-extended memory” property to its original definition. The approach can also be considered as a limit cycle construction technique for discrete dynamical systems.

WALSH TRANSFORM AS METHOD OF MIMO SYSTEMS IDENTIFICATION

Andrzej Żak

The paper presents method of MIMO system identification using Walsh transform. Paper includes description of mathematical basis of Walsh Transform. At the end of paper the results of research of identification for example multi input multi output object were presented.

A COMPUTER ORIENTED ALGORITHM FOR ANALYZING LIMIT CYCLES IN DISCRETE CONTROL SYSTEMS

M. Utrilla-Manso, R. Jiménez-Martínez, R. Mallol-Poyato, J. Sánchez-Golmayo and F. López-Ferreras

In this paper a new and fast algorithm for characterizing the behaviour of zero-input limit cycles that can appear in digital control systems when finite precision computer is used. This proposed algorithm suggests a practical approach to determine the impact of these parasitic oscillations against difficult theoretical solutions limited to simple systems and very conservatives in some cases. This algorithm is applicable to any kind of discrete system described by its difference equations and quantized by any quantization scheme and supply practical results in considerable less time that other exhaustive formulations. Some tables show the feasibility of the algorithm compared with exhaustive searches and theoretical calculations to characterize the limit cycles and its applicability for any kind of discrete system like different digital filters and digital control systems where different controllers are applied.

WAVELET TRANSFORM MOMENTS FOR FEATURE EXTRACTION FROM TEMPORAL SIGNALS

Ignacio Rodriguez Carreño and Marko Vuskovic

A new feature extraction method based on five moments applied to three wavelet transform sequences has been proposed and used in classification of prehensile surface EMG patterns. The new method has essentially ex-tended the Englehart's discrete wavelet transform and wavelet packet transform by introducing more efficient feature reduction method that also offered better generalization. The approaches were empirically evaluated on the same set of signals recorded from two real subjects, and by using the same classifier, which was the Vapnik's support vector machine.

A FAULT-TOLERANT DISTRIBUTED DATA FLOW ARCHITECTURE FOR REAL-TIME DECENTRALIZED CONTROL

Salvador Fallorina, Paul Thienphrapa, Rodrigo Luna, Vu Khuong, Helen Boussalis, Charles Liu, Jane Dong, Khosrow Rad and Wing Ho

Complex control-oriented structures are inherently multiple input, multiple output systems whose complexities increase significantly with each additional parameter. When precision performance in both space and time is required, these types of applications can be described as real-time systems that demand substantial amounts of computational power in order to function properly. The failure of a subsystem can be viewed as the extreme case of a non-real-time response, so the ability of a system to recognize and recover from faults, and continue operating in at least some degraded mode, is of crucial importance. Furthermore, the issue of fault-tolerance naturally arises because real-time control systems are often placed in mission-critical contexts. Decentralized control techniques, in which multiple lower-order controllers replace a monolithic controller, provide a framework for embedded parallel computing to facilitate the fault-tolerance and high performance of a sophisticated control system. This paper introduces a fault-tolerant approach to the handling of data flows in multiprocessor environments that are reminiscent of control systems. The design is described in detail and compared against a typical master-slave configuration. A distributed data flow architecture embraces tolerance to processor failures while satisfying real-time constraints, justifying its use over conventional methods. Both master-slave and distributed data flow designs have been studied with regards to a physical control-intensive system; the conclusions indicate a sound design and encourage the further division of computational responsibilities in order to promote fault-tolerance in embedded control processing systems.

WIND TURBINE ROTOR ACCELERATION: IDENTIFICATION USING GAUSSIAN REGRESSION

W. E. Leithead, Yunong Zhang and Kian Seng Neo

Gaussian processes prior model methods for data analysis are applied to wind turbine time series data to identify both rotor speed and rotor acceleration from a poor measurement of rotor speed. In so doing, two issues are addressed. Firstly, the rotor speed is extracted from a combined rotor speed and generator speed measurement. A novel adaptation of Gaussian process regression based on two independent processes rather than a single process is presented. Secondly, efficient algorithms for the manipulation of large matrices are required. The Toeplitz nature of the matrices is exploited to derive novel fast algorithms for the Gaussian process methodology that are memory efficient.

ROBUST STABILITY ANALYSIS OF SINGULARLY PERTURBED MAGNETIC SUSPENSION SYSTEMS

Nan-Chyuan Tsai and Chien-Ting Chen

For a singularly perturbed magnetic suspension system, two kinds of state feedback controllers are synthesized to account for the inherent instability of the open-loop plant with two-time-scale properties. Kharitonov polynomials, extremal vertex and uncertain Nyquist plot are employed to examine the maximum tolerance against system parameters uncertainties such that the stability of the closed-loop system is still retained. Experimental simulations are reported to illustrate the robustness of designed controllers both in stability and performance. At last, Interlacing Theorem is introduced to analyze the stability of uncertain suspension systems via the characteristic interval polynomials. It is found that identical results are obtained, in comparison with extremal vertex approach.

A PARAMETERIZED POLYHEDRA APPROACH FOR THE EXPLICIT ROBUST MODEL PREDICTIVE CONTROL

Sorin Olaru and Didier Dumur

The paper considers the discrete-time linear time-invariant systems affected by input disturbances. The goal is to construct the robust model predictive control (RMPC) law taking into account the constraints existence from the design stage. The explicit formulation of the controller is found by exploiting the fact that the optimum of a min-max multi-parametric program is placed on the parameterized vertices of a parameterized polyhedron. As these vertices have specific validity domains, the control law has the form of a piecewise linear function of the current state. Its evaluation replaces the time-consuming on-line optimization problems.

SYSTEM OF MEASURE AND REPRESENTATION OF ELECTROMAGNETIC EMISSIONS

Rafael Herradón Díez, Juan Fernandez-Corugedo, Julia Galiano Adán and Florentino Jimenez

Due to the vertiginous increase of the electromagnetic emissions of the modern communication systems and to the affected population's concern, it becomes necessary a more exhaustive control of this type of contamination. There are available different measure devices that make a characterization of the received emissions, but most doesn't discriminate against the contributions made by the different diffusion systems, So you cannot identify the systems responsible for the excessive measured levels. Of all the studied devices, the one that better it fulfills the expectations it is the spectrum analyzer, that it allows to characterize the received emissions, separating the different bands of the spectrum, and isolating the frequencies that don't perform the existent norm. To control the spectrum analyzer a program has been created that makes the measures according to the procedure exposed in the norm. Finally, a software has been developed that it processes and it represents the results of the measures, and it can calculate the theoretical securities of emissions in a location.

EXTRAPOLATION WITH A SELF-ORGANISING LOCALLY INTERPOLATING MAP - Controlling nonlinear Processes with ambiguous inverse Behaviour

Helge Hülsen and Sergej Fatikow

Besides theirs typical classification task, Self-Organizing Maps (SOM) can be used to approximate input-output relations. They provide an economic way of storing the essence of past data representing the relation into support vectors. In this paper the SOLIM algorithm (Self-Organising Locally Interpolating Map) is re-viewed and an extrapolation method is introduced. This framework allows finding one inverse of a nonlin-ear many-to-one mapping by exploiting the inherent neighbourhood criteria of the SOM part. Simulations show that the performance of the mapping including the extrapolation is comparable to other algorithms.

HYBRID ALGORITHMS FOR THE PARAMETER ESTIMATE USING FAULT DETECTION, AND REACHING CAPACITIES

Ryadh H. Mokhneche, Hichem Maaref and Vincent Vigneron

The nonstationary systems parametric estimate requires the continuation of its parameters which vary abruptly at unknown random moments. These are the abrupt parametric variations which were considered in this work to be managed like "faults". The considered signals here are nonstationary and are characterized by time variable parameters. The estimate of these parameters requires the choice of an algorithm having the capacity to continue their evolution. The various hybrid adaptive estimate methods showed that these capacities can be reached by a compensation of a gain and its update in online. In this paper, a method of estimate is proposed, based on the fault detection. The general algorithm implemented gives place to several methods which will be detailed. Experimental tests of some methods on a second order autoregressive synthesis signal are carried out and then commented.

COMPARATIVE PERFORMANCE OF INTELLIGENT IDENTIFICATION AND CONTROL ALGORITHMS FOR A FLEXIBLE BEAM VIBRATION

M. A. Hossain, A. A. Madkour and K. P. Dahal

This research presents an investigation into the relative real-time performance in implementing intelligent system identification and control algorithms. Several approaches for on-line system identification and control are explored and evaluated to demonstrate the merits in implementing the algorithms in real-time for similar level of error convergence. Active vibration control (AVC) of a flexible beam system is considered as a platform for the investigation. The AVC system is designed using three different on-line identification approaches, which include (a) genetic algorithms (GAs) (b) adaptive neuro-fuzzy inference system (ANFIS) and (c) recursive least square (RLS) estimation. These algorithms are used to estimate a linear discrete model of the system. Based on these algorithms, different approaches of the AVC system are implemented, tested and validated to evaluate the relative merits of the algorithms. Finally, a comparative performance of the error convergence and real-time performance in implementing the identification and control algorithms is presented and discussed through a set of experiments.

MERGING OF DATA KNOWLEDGE IN BAYESIAN ESTIMATION

Jan Kracík and Miroslav Kárný

Efficient multiple participant decision-making relies on cooperation of participants. Partially, it is reached by sharing knowledge. A specific but important case of this type is addressed here. Essentially, a participant passes to its partner distribution on common data and partner uses it for correcting its Bayesian parameter estimate.

IMPROVEMENT ON THE POLE-PLACEMENT CONTROL SCHEME BY USING GENERALIZED SAMPLED-DATA HOLD FUNCTIONS

David J. Donaire, Rafael Bárcena and Koldo Basterretxea

This paper studies the benefits the use of GSHF can afford to the pole-placement control scheme. The GSHF makes possible to locate the zeros of the discretized plant arbitrarily in the Z plane. This property can be taken advantage of to improve the performance of the pole-placement control. In this article a new design method is suggested and a simulations-based application example is carried out. In the application example the improvements this method involves with respect to the classical design method are noticed.

SYSTEMATIC APPROACH TO MODEL-BASED DATA SURVEY

Ari Isokangas, Mika Ruusunen and Kauko Leiviskä

A framework for surveying multivariate process data is presented. Systematic procedure utilises linear model candidates constructed in sliding data windows of varying length, to determine the usefulness of data segments for process identification. The discussed survey approach was applied to an industrial wood debarking data, enabling the study of process variables and conditions affecting the wood losses. In addition, main process interactions and delays were easily discovered from the structures of the interpretable linear model candidates. The analysis can thus provide valuable information also for process modelling and control.

MODELING OF MOTOR NEURONAL STRUCTURES VIA TRANSCRANIAL MAGNETIC STIMULATION

Giuseppe d’Aloja, Paolo Lino, Bruno Maione and Alessandro Rizzo

Transcranial Magnetic Stimulation (TMS) of human motor area can evoke different biological waves in the epidural space of patients. These waves can evoke different muscle responses according to different types and amplitudes of stimuli. In this paper we analyze the different types of epidural waves and we propose a neuronal model for the biological structures involved in the experiments.

OPTIMAL CONTROL APPLIED TO OPTIMIZATION OF MOBILE SWITCHING SURFACES PART I: ALGORITHM

Jean-Claude Jolly, Céline Quémard and Jean-Louis Ferrier

Following Boccadoro et al., 2004, we consider hybrid dynamical systems with parameterized switching surfaces. The goal is to optimize the choice of parameters in relation with a criterion. In an optimal control framework we deepen and generalize results of these authors. An original application of the dynamic programming principle enables us to recover known expressions for Hamitonian and costate jumps that yield to an algorithm of resolution. What is interesting in the studied problem is that the found algorithm, usually not totally explicit, can here be specified up to obtain an efficient one. Ideas of some new or classical future applications are given.

DECOMPOSITIONS OF HIERARCHICAL STATE ESTIMATION STRUCTURES - Problems and Strategies

Rogério Bastos Quirino and Celso Pascoli Bottura

This study has three main objectives. First, to point and discuss the principal features, advantages, and limitations of didtributed state estimators. Second, to analyze structures and methodologies related to the distributed state estimation problem, with emphasis on the heterarchical one. Finally, to delineate some prospects for future investigations.

IMPROVED STABLE FEEDBACK ANC SYSTEM WITH DYNAMIC SECONDARY PATH MODELING

Rogelio Bustamante-Bello, Héctor Pérez-Meana and Bohumil Psenicka

This paper presents the development and DSP implementation of a stable ANC feedback system with on-line secondary path modelling, using the Normalized Filtered-X Least Mean Square with Noise Addition algorithm (NFXLMS-NA), for acoustic noise cancellation. In this paper, the feedforward and the feedback ANC systems are described briefly; the basic of the FXLMS algorithm and its structure is discussed and the new NFXLMS-NA algorithm is presented. The ANC system developed includes a conventional noise predictor, a primary adaptive filter, a subsystem for dynamic secondary path modelling and the addition of white noise signal in the FXLMS algorithm in a novel structure looking for stability into the system. The system was developed for cancelling quasi-periodic acoustic noise; some experimental results for narrow-band signal are included in order to show the desirable feature (stability) of the system. Proposed system was implemented using a TMS320C30 evaluation module from TI. Finally, the paper includes the block diagram of the ANC system, the structure of the program used in the implementation and some photographs of the practical scheme and the equipment used in the tests.

EXTRACTION OF OBJECTS AND PAGE SEGMENTATION OF COMPOSITE DOCUMENTS WITH NON-UNIFORM BACKGROUND

Yasser Alginahi, Maher Sid-Ahmed and Majid Ahmadi

In designing page segmentation systems for documents with complex background and poor illumination, separating the background from the objects (text and images) is very crucial for the success of such system. The new local based neural binarization technique developed by the authors will be used to extract the objects from document images with complex backgrounds. This algorithm uses statistical and textural feature measures to obtain a feature vector for each pixel from a window of size (2n+1)x(2n+1) , where n>=1. These features provide a local understanding of pixels from their neighbourhoods making it easier to classify each pixel into its proper class. A Multi-Layer Perceptron Neural Network (MLP NN) is then used to classify each pixel in the image. The results of thresholding are then passed to a block segmentation stage. The block segmentation technique developed is a feature-based method that uses a Neural Network classifier to automatically segment and classify the image contents into text and halftone images. The results of page segmentation are then ready to be passed into an OCR system that will convert the text image into a format the can be stored and modified.

TESTBED EVALUATION OF NETWORKED CONTROL SYSTEMS

George Hassapis, Spyridon Geronatsios and John Grigoriadis

This work addresses the issue of performance evaluation of advanced control algorithms which are going to be implemented on scalable industrial computer networks. The basic characteristic of such an implementation is that information concerning measurements from sensors, commands to actuators and reference inputs is exchanged between the plant and the control system over a real-time communication network. The need to evaluate the performance of an algorithm implementation before commissioning it derives from the fact that the network-induced delay in the exchange of the sensor-to-controller and controller-to-actuator data and the possibility of loss of a data package during transmission may affect the algorithm performance. One way of assessing this performance is by emulating the operation of the algorithm on a test-bed. As test-bed is defined the facility that consists of a computer-based simulation of the plant which is linked to the real communication network and actual control devices on which the algorithm will be implemented. As there are many proprietary and open communication network protocols and standards, unavoidably the test-bed has to be constructed for a specific protocol and standard. In this work a test-bed based on the Profibus standard and its FMS protocol has been realized. The purpose is the evaluation of a control algorithm which will run on one or more controllers that will be inserted in an already operating networked control system. In order to demonstrate the way of using a test-bed for evaluating the performance of a control algorithm the study of the LQC control of a cement milling circuit is presented.

COLOR IMAGE SEGMENTATION BY GRAVITATIONAL CLUSTERING IN COLOR SPACE USING NEIGHBOR-RELATIONSHIP

Hwang-soo Kim and Hwajeong Lee

In this paper, we propose a color image segmentation method based on gravitational clustering using neighbor relations in the spatial domain and distance information in RGB space among pixels. Most clustering-based segmentation algorithms use only color space distances after pixels are mapped from the spatial domain to color space, ignoring their neighbor relations; but we use both information. We use gravitational clustering, which imitates the Law of Gravity, and the gravitational force is applied only to neighboring clusters. The results show that the proposed method is effective in finding exact boundaries of regions

OPTIMAL CONTROL APPLIED TO OPTIMIZATION OF MOBILE SWITCHING SURFACES PART II : APPLICATIONS

Céline Quémard and Jean-Claude Jolly

An application of the other paper submitted "Optimal Control of Mobile or not Swithching surfaces in Hybrid Dynamical Systems" is presented. It concerned a thermostat with anticipative resistance. We optimize the adjustment of thermostat tresholds to control at best the temperature of the room.

A MODEL BASED HYBRID NUMERICAL CONTROL ALGORITHM FOR THE CONTINUOUS DRYING OF A THICK WEB IN AN INFRARED DRYER

Normand Thérien, Arthur Broadbent and Sergio Pérez

Experimental results from the transient drying of sheets of polyester in an infrared (IR) dryer were used to derive a performance model. Separate drying experiments were done using sheets of material of various densities and thicknesses. The formulations expressed the core temperature of the web to the surface temperature of the web as a function of the residency time in the dryer and the electric power used. Also, a relationship between the time duration required to achieve a given core temperature of the web as a function of the electric power was derived. These relationships were used to derive an hybrid numerical control algorithm using feed forward and feedback actions to control the core humidity of the web at the outlet of the dryer.

REAL-TIME MODELLING OF WOOD DRYING SYSTEMS - Learning from Experiment and Theory

Stanislaw Tarasiewicz and Belkacem Kada

Predictive control in a wood drying systems is still at an early stage, because of the difficulties with the estimating a temporal moisture distribution for the whole dried lumber. Therefore, based on the dry and wet-bulb temperatures as the state variables the temporal moisture distribution in kiln-dried lumber is determined from numerical solutions of mathematical model for the wood drying systems. This computer model is represented by a set of several partial nonlinear differential equations coupled with the operating functions and a set of several calculating algorithms. The accuracy of the model solutions in a real-time calculation is evaluated by the on-line identification of the operating functions that represent both the system parameters (heat transfer coefficients, thermal conductivity, heat capacity etc) and selected state variables (air temperature, humidity, velocity etc).

CONFIDENCE BASED ESTIMATION AND DETERIORATION INDICATION OF ON-LINE MEASUREMENT

Jari Näsi and Aki Sorsa

In an industrial process, the accuracy and reliability of process creates basics for control system and ultimately to product uniformity. Measurement results, whether from fast on-line sensors or from sample-based laboratory analyses, is the key for selecting the method for process control and analysis. Intelligent and advanced control methods, exploiting measurements, are of no benefit if the measurements cannot be trusted. This paper presents an estimation method for combining real-time redundant signals, consisting of sensor data, and analytical measurements. The validation of on-line measurement uses less frequently updated but more accurate information to validate frequently updated but less accurate on-line measurements. An estimate of the measured variable is obtained as a weighted average of the on-line measurements and laboratory analyses. The weighting coefficients are recursively updated in real time when new analysis and measurement results are available. The calculation of optimal estimate can be used in several industrial applications for more precise process control. In addition, pre-processed data is used to calculate a “need for maintenance indicator” to warn the operator for sensor breakdowns, wearing or deterioration and detect calibration needs. The operator’s workload is reduced in problematic situations where measurement and validation signals are not convergent, by offering calculated best estimation.

CONTROL SYSTEM INTERFACE OF SCANNING SONAR FOR MOBILE ROBOTS

Sv. Noykov and O. Manolov

In this work, a simple, low-cost and reliable electronic module for coupling of an ultrasonic range-finder with a mobile robot’s microprocessor system is presented. A software filter for correct reading of the ultrasonic data is presented as well. Due to the software filter, a shielding of robot and sonar’s electronic modules is not required. In this way compactness and low price of the device construction were achieved.

INVARIANT SIGNAL RECOGNITION IN NOISE ENVIROMENT

Riad Taha Al-Kasasbeh

Recognition of signals is considered at the presence of noise. The problem is actual for the speech signals recognition, acoustic diagnostics of mechanisms. The model of recognized signal contains a priori unknown parameter - the relation a signal / noise. For considered model the approach to construction of the signal description, which is invariant to a level of the noise is offered. Efficiency of invariant signals recognition is analyzed.

Combining Multiple Pairwise Neural Networks Classifiers: A Comparative Study

Olivier Lezoray and Hubert Cardot

Classifier combination constitutes an interesting approach when solving multiclass classification problems. We review standard methods used to decode the decomposition generated by a one-against-one approach. New decoding methods are proposed and are compared to standard methods. A stacking decoding is also proposed and consists in replacing the whole decoding by a trainable classifier to arbiter among the conflicting predictions of the binary classifiers. Substantial gain is obtained on all datasets used in the experiments.

Neural Network Modeling for ALSTOM Gasifier

Armando Rivadeneyra Bardales, Danilo Soares Barboza, William Ipanaqué and Martin Flores

Neural Network Model Based Predictive Control (MPC) has become a good choice of control strategy in many cases especially in the process industry because it could face non linearities and crosscoupling variables [6], being modeling the first step to achieve this end. The model of a gasifier, provided by ALSTOM Power Technology Centre, is of an industrial standard and has been validated against a set of real data from test facilities. This makes the challenge all the more relevant to practicing engineers. The paper sets out the specifications and describes the design and performance of neural netwoks modeling. This paper presents a neural network approach to model the ALSTOM Benchmark Challenge gasifier. This is a complex non-linear process, with a high degree of cross coupling of the variables, manual control is difficult.

Grinding Forces Prediction Based Upon Experimental Design and Neural Network Models

Ridha Amamou, Nabil Ben Fredj and Farhat Fnaiech

The results presented are related to the prediction of the specific grinding force components. The main problems associated with the prediction capability of empirical models developed using the design of experiment (DOE) method are given. In this study an approach suggesting the combination of DOE method and artificial neural network (ANN) is developed. The inputs of the developed ANNs were selected among the factors and interaction between factors of the DOE depending on their significance at different confidence levels expressed by the value of %. Results have shown particularly, the existence of a critical input set which improves the learning ability of the constructed ANNs. The built ANNs using these critical sets have shown low deviation from the training data and an acceptable deviation from the testing data. A high prediction accuracy of these ANNs was tested between models constructed using the developed approach and models developed by previous investigations.

A Support System for fisheries based on Neural Networks

Alfonso Iglesias, Bernardino Arcay, Alejandra Rodríguez and Manuel Cotos

This paper presents the foundations of a decision support system for the localisation of fisheries based on AI techniques. The purpose of such a system is to reduce the costs of fishing fleets without endangering the sustainable development of the natural resources. Our data sources are satellite images (OrbView-2, Series NOAA, Topex/Poseidon), as well as real catch data obtained from the fishing log of a pilot boat. We have compared neural networks, ANFIS, and functional networks, and we have exported the results to a SIG. The best results were obtained for a perceptron trained with the Backpropagation method.

Hybrid Strategy for Automatic Stellar Classification

Alejandra Rodriguez, Carlos Dafonte, Bernardino Arcay, Iciar Carricajo and Minia Manteiga

This paper describes an hybrid approach to the unattended classification of optical spectra of stars. The classification of stars in the standard MK system constitutes an important problem in the Astrophysics area, since it helps to carry out proper stellar evolution studies. Manual methods, based on the visual study of stellar spectra, have been frequently and successfully used by researchers for many years, but they are no longer viable because of the spectacular advance of the objects collection technologies, which gather a huge amount of spectral data in a relatively short time. We therefore propose a cooperative system that is capable of classifying stars automatically and efficiently, by applying to each spectrum the most appropriate method or combined methods, which guarantees a reliable, consistent and adapted classification. Our final objective is the integration of several artificial intelligence techniques in a unique hybrid system.

A New Training Algorithm for Neuro-Fuzzy Networks

Stefan Jakubek and Nikolaus Keuth

In this paper a new iterative construction algorithm for local model networks is presented. The algorithm is focussed on building models with sparsely distributed data as they occur in engine optimization processes. The validity function of each local model is fitted to the available data using statistical criteria along with regularisation and thus allowing an arbitrary orientation and extent in the input space. Local models are consecutively placed into those regions of the input space where the model error is still large thus guaranteeing maximal improvement through each new local model. The orientation and extent of each validity function is also adapted to the available training data such that the determination of the local regression parameters is a well posed problem. The regularisation of the model can be controlled in a distinct manner using only two user-defined parameters. In order to assess the quality of the obtained model the algorithm also provides accurate model statistics. Different examples illustrate the efficiency of the proposed algorithm. One illustrative example shows how local models are adapted to the shape of the target function in the presence of noise. A second example shows results obtained with measurement databases of IC-engines. The last example features the identification of a nonlinear dynamic process.

Evolutionary Techniques for Neural Network Optimization

Eva Volná

The idea of evolving artificial networks by evolutionary algorithms is based on a powerful metaphor: the evolution of the human brain. The application of evolutionary algorithms to neural network optimization is an active field of study. The success and speed of training of neural network is based on the initial parameter settings, such as architecture, initial weights, learning rates, and others. A lot of research is being done on how to find the optimal network architecture and parameter settings given the problem it has to learn. One possible solution is use of evolutionary algorithms to neural network optimization systems. We can distinguish two separate issues for it: on the one hand weight training, and on the other hand architecture optimization. Next, we will focus on the architecture optimization and especially on the comparison of different strategies of neural network architecture encoding for the purchase of the evolutionary algorithm.

Genetic Tuner for Image Classification with Robotic Applications

Shahram Jafari and Ray Jarvis

This paper introduces two genetic tuners implemented for optimizing the image segmentation and scene analysis prior to the grasping of the objects, to realize a concrete working robot named COERSU. Firstly, a robust tuner is presented to optimize the early visual processing (image segmentation and edge detection) based on genetic algorithms (GA). Then, different architectures of the adaptive neuro-fuzzy inference system (ANFIS), multi-layer perceptron (MLP) and K-nearest neighborhood (KNN) classifiers are compared to perform scene analysis and object recognition. Following on, the MLP classifier is chosen due to its accuracy and flexibility to be tuned by genetic algorithm (GA). The real-time experiments (after tuning) show that the performance of the genetically tuned MLP classifier is improved in terms of accuracy due to this hybridization. Finally, snapshots of the experimental results from COERSU in a table-top scenario to manipulate some soft objects (e.g. fruit/egg) are provided to validate the methods.

Can a Fuzzy Rule Extraction Find an Extremely Tiny non-Self Region?

Akira Imada

This paper reports one snapshot of our on-going experiments in which a common target we call a-tiny-island-in-a-huge-lake is explored with different methods ranging from a data-mining technique to an artificial immune system. Our implicit interest is a network intrusion detection where we usually do not know what does an illegal transaction pattern look like until it completed intrusion when it was too late. Hence our first interest is (i) if it is possible to train the intrusion detection system only using legal patterns. From this context we assume data floating in the lake are normal while ones found on the {island} is abnormal. Our second concern is then (ii) to study the limit of the size of the detectable area, that is, when we decrease the size of the island shrinking to zero until which size can the detector detect it. In this paper a fuzzy rule extraction implemented by a neural network architecture is employed.for the purpose,

An Energy Efficient Multiagent Middleware for Physical Complex System

Jean-Paul Jamont, Michel Occello and André Lagrèze

Open physical complex artificial systems involve wireless autonomous entities submitted to strength constrainted energetic policies. The features of these systems naturally leads to apply multiagent techniques to ensure both the autonomy of entities and the best organization of the whole system. We propose a multiagent approach for wireless communication management and functional integrity maintaining for such physical multiagent systems using self-organization mechanisms. We show the performances of this multiagent approach by comparison to usual communication management based on wireless protocols used in this area. The genericity of our contribution is highlighted by the proposition of a middleware layer integrated in agent. We give an insight to a real world application of intrumentation of an underground instrumentation using the middleware.

An Architecture-Altering and Training Methodology for Neural Logic Networks: Application in the banking sector

Athanasios Tsakonas and Georgios Dounias

Artificial neural networks have been universally acknowledged for their ability on constructing forecasting and classifying systems. Among their desirable features, it has always been the interpretation of their structure, aiming to provide further knowledge for the domain experts. A number of methodologies have been developed for this reason. One such paradigm is the neural logic networks concept. Neural logic networks have been especially designed in order to enable the interpretation of their structure into a number of simple logical rules and they can be seen as a network representation of a logical rule base. Although powerful by their definition in this context, neural logic networks have performed poorly when used in approaches that required training from data. Standard training methods, such as the back-propagation, require the network’s synapse weight altering, which destroys the network’s interpretability. The methodology in this paper overcomes these problems and proposes an architecture-altering technique, which enables the production of highly antagonistic solutions while preserving any weight-related information. The implementation involves genetic programming using a grammar-guided training approach, in order to provide arbitrarily large and connected neural logic networks. The methodology is tested in a problem from the banking sector with encouraging results.

Fuzzy Controller for Flatness Based on Neural Network Pattern-recognition

Jian-chang Liu, Zhu Wang

A pattern-recognition method for flatness defect based on CMAC neural network is proposed, and a flatness fuzzy controller based on the pattern-recognition results is designed in this paper. Pattern-recognition and controller are designed into a single unit, in which CMAC recognizes the membership grade relative to six basic modes of common flatness defect and realizes the seeking function of the membership grade as the forepiece of the fuzzy controller for flatness directly. Through analyzing the characteristics of the flatness defect, the fuzzy set is defined reasonably, which has greatly reduced the calculation amount of fuzzy reasoning. The result of simulation shows that the pattern -recognition method of flatness offers high recognizing precision, the designed fuzzy controller for flatness can control the flatness defect to expected goal fleetly and the performance of flatness control is fine.

Proposal Model for Stamping Application Using Artificial Neural Networks System

Noureddine Ben Yahia, Sabeur Abid and Ali Zghal

In this research, the approaches of feature stamping design and Artificial Neural Networks (ANN) are combined to automate the process planning task and to generate process groups for set-ups. The model created in Computer Aided Process Planning (CAPP) system can provides different process using ANN for cylindrical parts. This model is composed by three principal modules, the first relates to geometrical in 3D modeling, the second treats calculations of the stamping process parameters and the third module proposes the processes of obtaining a final part using ANN system. The development of this system is based on the experiments and the knowledge to make specialists in this field. Indeed in this work we started with a theoretical study concerning the influence of the parameters of stamping and the causes of the principal defects of an operation of working of the cylindrical parts and the proposal for several typical examples of processes which are validated with industrialists. In this work we focus only in ANN structure for this application, what is Input? What is output ? to give industrial solution.

Sub Auditory Communication and Facial EMG

Sanjay Kumar, Dinesh Kant Kumar and Melaku Alemu

Aavailability of speech related information in the facial EMG is discussed. The primary objective of this preliminary work is to investigate the use of facial EMG as a voiceless communication medium. Subjects were asked to utter the five English vowels with no acoustic output (sub-auditory). Three independent EMG signals were acquired from three facial muscles as sub-auditory EMG activations. In order to classify and recognize each vowel based on EMG, RMS (Root Mean Square) of the recorded signals were estimated and used as parametric inputs to a neural network.

Biometrics Identification Based on Visual Hand Movements Using Wavelet Transform

Sanjay kumar, Dinesh Kant Kumar and Neil Mclachalan

This work presents a novel technique of biometric identification based on the temporal history templates (THTs) of visual hand movements. The technique uses view-based approach for representation of hand movements, and uses a cumulative image-difference technique where the time between the sequences of images is implicitly captured in the representation of action. The low level representation of the action collapses the temporal structure of the motion from the video sequences of the hand movements while removing any static content from the video sequences to generate temporal history templates (THTs) of the hand movement. THTs of different individuals present distinctive 2-D motion patterns, where each pixel describes the function of temporal history of motion in that sequence. This THT are further sub-divided into four sub- images an average and three detailed images using multi resolution wavelet transforms. The approximate wavelet sub-image is considered as the feature for recognition. The recognition criterion is established using KNN nearest neighbor technique using Mahalanobis distance. The accuracy of accepting an enrolled subject (AAES %) and accuracy of rejecting an imposter (ARI %) are the indicators of identification performance of the technique. The experimental results from 5-different individual indicate that the THT based technique achieves high identification rate when subject specific movements are assigned to the subjects during enrolment.

Intrinsic Classification of Single Particle Images by Spectral Clustering

Yutaka Ueno, Masaki Kawata and Shinji Umeyama

An application of spectral clustering to single particle analysis of a biological molecule is described. Using similarity scores for the given data set, clustering was performed in a factor space made by the eigenvector of the normalized similarity matrix. Image data was thus classified by means of information intrinsic to the ensemble of given data. The method was tested on a simulated transmission electron microscopy image and a real image data set of 70S ribosome. The average images of clusters were obtained by iterative alignment, which successfully represented characteristic views of the target molecules. Comparisons with traditional methods and techniques in practical implementation are discussed.

Characterising Evoked Potential Signals Using Wavelet Transform Singularity Detection

Conor McCooey and Dinesh Kant Kumar

A new method of viewing evoked potential data is described. This method, called the peak detection method, is based on singularity detection using the discrete wavelet transform. The peaks and troughs of an EEG recording are identified and characterised using the algorithms of this method, resulting in a linear decomposition of the data into sets of individual peaks. Then, by classifying the peaks in terms of latency (time), magnitude (voltage potential) and width (scale), the locations of higher concentrations of similar peaks are identified. These are grouped and sub-averaged, yielding sets of sub-averaged peaks that characterise the shape and give a measure of the repeatability of particular sub-averages within the visual evoked potential ensemble average signal. Linearity is maintained allowing the ensemble average to be generated by adding the sub-averaged peaks together. This method highlights evoked potential features that may be obscured or cancelled out with standard ensemble averaging. This is demonstrated for visual evoked potential data taken from a single subject.

Source Extraction Representing Thumb and Little Finger Cortical Networks and Intra-Hand-Area Evoked Synchrony

Franca Tecchio, Sara Graziadio, Giulia Barbati, Roberto Sigismondi, Filippo Zappasodi, Camillo Porcaro, Giancarlo Valente and Marco Balsi

A novel cerebral source extraction method is proposed (Functional Source Separation, FSS) starting from extra-cephalic magnetoencephalographic (MEG) signals in humans, based on source functional reactivity to the external stimulation. Their activity is obtained all along the rest state and during proc-essing of a simple separate sensory stimuli of thumb or little finger. This method provides cerebral sources describing thumb and little finger primary representations, as demonstrated by their higher responsiveness to the corre-sponding finger stimuli and their different positions consistent with the homun-cular organization. A dynamical index describing intra-regional synchrony was introduced, which showed higher levels when stimulating the thumb with re-spect to stimulation of the little finger, selectively in the gamma band. This in-dicates that the stimulation of a functionally prevalent finger (thumb) activates a cortical network more synchronized in the gamma band than a non-prevalent one (little finger).

Foot Pathologies Classification From Pressure Distribution Over The Foot Plant

José García-Hernández, Roberto Paredes, David Garrido and Carlos Soler

Nowadays, approximately 35\% of the population suffers problems in the feet. In many cases, some foot pathologies are treated by means of especial insoles. These insoles require customisation for the success of their treatment. However, there is a great lack of knowledge in this sector. As an initial step it is essential to develop a new objective tool capable to classify among the different foot pathologies. In this paper we present a system to classify different foot pathologies which uses the Nearest Neighbor (NN) classification rule based on weighted metrics and prototypes selection. The system uses as input data the pressure distribution over the foot plant.

Efficient Gridding of Real Microarray Images

Giuseppe Lipori

DNA microarrays technology is very recent and rapidly evolving. At present, it is widely used in the analysis of gene expression. The interpretation of the data crucially depends on the accuracy of the localization of the circular spots, which are placed in rectangular grids. The problem is complicated by the presence of many local deformations of the grid, by the high variability in luminance of the spots, by noise and other disturbances due to the biological nature of the experiments. In this paper we propose an automatic method for the gridding of real microarrays that takes into account most of the open problems by exploiting a recently introduced image transform, the Orientation Matching Transform, which enhances circular patterns of a specific size.

SVM Classification of Sparse Set of 1:1 Ventricular and Supraventricular Tachycardia

Mario de-Prado-Cumplido, Ángel Arenal-Maíz, Mercedes Ortiz-Patón and Antonio Artés-Rodríguez

Inappropriate classification of Supraventricular Tachycardias with 1:1 atriovetricular conduction is a mayor issue in dual-chamber implantable cardioverter defribillators. In order to distinguish Supraventricular from Ventricular Tachycardias a new methodology is proposed. The tachyarrhythmia episodes ECGs are characterized into feature vectors, which are then classified using a Support Vector Machine. The best features of the vectors are selected by means of several Feature Selection methods. The performance of the algorithm overcomes existing algorithms for implantable devices.

Optimizing ICA Using Prior Information

Giancarlo Valente, Giuseppe Filosa, Federico De Martino, Elia Formisano and Marco Balsi

In this work we introduce a novel algorithm for Independent Component Analysis (ICA) that takes available prior information on the sources into account. This prior information is included in the form of a ?weak? constraint and is exploited simultaneously with independence in order to separate the sources. Optimization is performed by means of Simulated Annealing. We show how it outperforms classical ICA algorithms in the case of low SNR. Moreover, additional prior information on the sources enforces the ordering of the components according to their significance.

Functional Constraints Added to an ICA Separating Algorithm: an Example on Magnetoencephalographic Signals

Marco Balsi, Giuseppe Filosa, Roberto Sigismondi, Giancarlo Valente, Giulia Barbati, Filippo Zappasodi, Sara Graziadio, Camillo Porcaro and Franca Tecchio

A constraint function expressing a priori information about the structure of data recorded in a MEG experiment is used to bias ICA towards a more realistic decomposition that effectively succeeds in separating physiologically significant activities that standard ICA fails to distinguish.

HRV Computation as Embedded Software

George Manis

The heart rate signal contains useful information about the condition of the human heart which cannot be extracted without the use of an information processing system. Various techniques for the analysis of the heart rate variability (HRV) have been proposed, derived from diverse scientific fields. In this paper we examine theoretically and experimentally the most commonly used algorithms as well as some other interesting approaches for the computation of heart rate variability from the point of view of the embedded software development. The selected algorithms are compared for their efficiency, the complexity, the size of the object code, the memory requirements, the power consumption, the real time response and the simplicity of their interfaces. Figures giving a rough image of the capability of each algorithm to classify the subjects into two distinct groups presenting high and low heart rate variability are also presented, using data acquired from young and elderly subjects.

Measuring the Emotional and Physiological Effects of Light and Colour on Space Users

Nadeen Abbas, Dinesh Kumar and Neil Mclachlan

Colour and light are known to have an impact on our health and wellbeing. While large resources are allocated for well designed buildings with the right choice of colours and lighting conditions, there is little scientific evidence that supports these choices. The aim of this research was to determine the impact of different colours and lighting conditions on people, using non-invasive means. Close correlations between skin conductance (SC), our emotions and our health are well reported in literature and hence these are expected to be good measures of the environmental conditions on people.

Stress and Heart Rate Variability

Prashant Suryanarayanan and Dinesh Kant Kumar

Do people get accustomed to stress? Do people react differently to a stressful environment when they have experienced the situation earlier? This paper reports research to determine whether the above is a myth. It reports the measurement of heart rate variability (HRV) of people when subjected to controlled stress conditions and these conditions are repeated over different days. The results indicate that while there is a large variability between different participants, there is a consistent reduction in the heart rate variability from the first experience of the stress condition to the subsequent applications of similar stress conditions.

Comparison of Time and Spectral Domain Features on Postural Signals Utilizing Neural Networks

Andreas Fey, David Sommer and Martin Golz

Human postural equilibrium is the result of complex control processes. Nevertheless these processes are taken for granted in our daily life, disturbance or degeneration of a single system involved in these processes leads to a variety of diseases, which pile up with age. Therefore, investigation of postural signals is the aim of many clinical and biophysical studies, in order to recognize diseases early and to improve the precision of diagnostics. In order to analyze posturographic signals we conducted a pilot study to measure body sway of nine healthy subjects during four trials with different acoustic and visual impairments, in order to detect their influence on stance. Ten time domain and five spectral domain feature extraction methods were applied on segmented raw data and classified by five different classification methods. The test errors were empirically minimized first by estimating best parameters for each feature extraction method, yielding to an optimal combination of feature extraction and classification methods. It turned out, that Burg autoregressive method of power spectral density estimation and Optimized Learning Vector Quantization was the best method combination. The classification task “no impairment” versus “visual impairment”, i.e. “eyes open” versus “eyes closed”, showed best discriminative performance indicated by mean test errors of 2.2%. The pilot study pointed out, that the established biosignal analysis system gained a high sensitivity on small postural influences.

Independent Component Analysis and Raman Microspectroscopy on Paraffinised Non Dewaxed Cutaneous Biopsies: A promising Methodology for Melanoma Early Diagnosis

Cyril Gobinet, Ali Tfayli, Olivier Piot, Valeriu Vrabie and Régis Huez

This paper deals with a promising methodology for melanoma early diagnosis. Raman spectroscopy is used to record vibrational information of paraffinised tumoral tissues. Independent Component Analysis (ICA) is performed on Raman spectra to numerically deparaffinise spectra. Resulting deparaffinised spectra are used to extract discriminant information specific to malignant and benign tumors. These spectral specificities can be employed as molecular descriptors of the type of pathology. A comparison with Principal Component Analysis (PCA) shows that ICA is more suited to process this kind of problem.

Microsleep Detection in Electrophysiological Signals

Martin Golz, David Sommer and Danilo Mandic

An adaptive biosignal analysis system for the detection of microsleep events is presented. The system was applied to the electroencephalogram and electrooculogram recorded of 23 young volunteers which performed monotonic overnight driving in our real car driving simulation laboratory. Clear microsleep and clear non-microsleep events were scored by experts evaluating video recordings independently of the biosignals to be analyzed. Besides the commonly used Periodogram method to estimate power spectral densities we utilized the recently established method of Delay Vector Variance. The so obtained feature set was used as input vectors of populations of Learning Vector Quantization networks which are evolved using Genetic Algorithms. This method is compared to Support Vector Machines which attained best performance. Fusion on feature level of all recorded signals and of both feature types led to empirical test errors down to 11.2 %. It is shown that the proposed methodology is not able to predict immediately oncoming events.

The Effect of White Noise and False Peak Detection on HRV Analysis

G. Manis, A. Alexandridi, S. Nikolopoulos and K. Davos

Heart rate variability (HRV) is an established measure for cardiac health. Its use is widespread and many methods have been developed for its analysis. Little emphasis, however, has been given to the specific influence of noise from the electrocardiogram (ECG) on the heart rate (HR) series. There are explicit factors of noise that have been extensively studied on the ECG and much work has been published on their limitation or elimination. Despite all these solutions, however, often noise does end up in the ECG and is inevitably included in the derived HR series. It is of interest to investigate how this influences subsequent HRV analysis. We propose that the noise into the resulting HR series: white noise and false peaks. In this paper, we demonstrate how these two scenarios affect the outcome of the HRV analysis.

The origin of artificial kinetic spectroscopy and its application

Ilya Fine

Intentionally creating local blood flow cessation at different body sites initiates the time dependent optical response. Specifically, applying circumferential over systolic pressure at the finger base creates a very stable post-occlusion optical signal (POS). Geometrically independent features of light absorption and scattering of the blood are extracted from POS by evaluating the parametric slope (PS). It was shown that PS, being determined from suitable pairs of wavelengths, correlates with arterial blood oxygen saturation. Other PS pairs exhibited a strong correlation with blood he-moglobin. The time dependent signal was simulated in vitro by using the red blood cells (RBC) aggregation process and was shown to resemble the main features of in vivo POS.

On Cluster Analysis Via Neuron Proximity in Monitored Self-Organizing Maps

Susana Vegas-Azcárate and Jorge Muruzábal

The topographic or self-organizing map (SOM) has become a widely-used mining tool to ascertain structure in the data. We are particularly interested in the application of the SOM neural structure as a clustering tool for biological signals. Traditionally, clusters in SOM structures have been detected using strategies that place the main emphasis on some graphical, neuronwise summary of pointer interdistances. While it is known that some care must be taken when fitting these maps to data, a complete methodology of topographic map formation is not available yet. In a companion paper, we have introduced an early-stopping criterion called UDL, and we have seen that it provides sensible density estimates for the unknown sampling ditribution (when such estimates are produced by the training algorithms). In this paper we review the more general approach based on neuron interdistances alone. Four different algorithms are studied using both artificial and real data. Our results suggest that all these training algorithms can indeed go a long way by restricting attention to neuron proximity, but the emerging maps must be monitored somehow.

Statistical Analysis of the Human EEG during RF Exposure from Mobile Phones: An Alternative Method to Analysis of the EEG in Frequency Bands

Howard D’Costa and Irena Cosic

This paper aims to describe a novel statistical approach to analysing the effects of radiofrequency (RF) exposures from mobile phones on the human EEG. In addition, the paper describes two limitations that may be encountered when using statistical methods to analyse the EEG in its frequency bands. The proposed method of analysis which is based on measures of central tendency introduces an approach whereby the recorded body of EEG data collected during trials can be effectively interpreted for spectral analysis at a higher resolution across the EEG spectrum. It is believed that the proposed statistical approach may be also useful in other studies investigating the effects of alternate forms of involuntary stimulus on the human EEG, such as electrical stimulus, light, and sound.

Cooperative Task Scheduling Among Time-Bounded Agents

Habiba Belleili, Maroua Bouzid and Mokhtar Sellami

The paper describes an approach to the cooperative execution of tasks that can benefit from the application of multiple methods. Agents have alternative methods for task solving, ranging from approximate ones to others that are more precise but also more resource (time) demanding. Tasks have temporal constraints (deadlines). Agents are resource-limited and operate under time constraints. Agents go into a negotiation process in order to choose a combination of methods which maximizes the utility of the result. Agents (with their chosen methods) will be used in sequential levels to progressively improve the quality of a task solution.

Cooperative Control Of A Robot Swarm With Network Communication Delay

Yechiel Crispin

A new method for the cooperative control of a system of multiple mobile robots with time delay in the network communication is presented. The network of mobile robots is modeled as a swarm of particles performing a directed random walk where the motion of the swarm is controlled by a central unit such as a robot leader. The collective motion of the robots is modeled by a system of stochastic delay-difference equations where the best solution found by the swarm is used as the network cooperative control signal. The method is applied to the solution of two problems. In the first problem, a group of autonomous underwater vehicles (AUVs) searches for the maximum depth in a two-dimensional domain. In the second problem, the group of AUVs searches for the minimum temperature in a three-dimensional domain. It is found that the search proceeds along Levy flights followed by sticking short random walks in the vicinity of the extremum points and that the cooperative control method is robust to time delays in network communication.

A Tiny Overview of Cfengine: Convergent Maintenance Agent

Mark Burgess

Cfengine is a distributed agent framework for performing policy-based network and system administration that is used on hundreds of thousands of Unix-like and Windows systems. This paper describes cfengine's stochastic approach to policy implementation using distributed agents. It builds on the notion of `convergent' statements, i.e. those which cause agents to gravitate towards an ideal configuration state, which is implied by policy specification. Cfengine's host classification model is briefly described and the model is compared to related work.

Introducing Simulation Middle Size: A New Soccer League to RoboCup

Mansour Jamzad and Mahdi Asadpour

There is a big difference between Simulation League and Middle Size League in RoboCup, that is, a program written for the former can not be easily transferred to the latter. Since the final aim is real robotics, in this paper we suggest a new match for RoboCup Soccer which can fill this gap. We designed and implemented a Modeled Robot Soccer Server most like Middle Size RoboCup that contains a Soccer Filed, Ball, Referee, etc. Each participated team should provide not only his robots controller program but also the hardware model of own robots in VRML97 format, using which our system will model a simulation version of them. As an implementation of our ideas, we modeled and simulated two teams of robots and present the set up method for the match between them. According to these setups, two teams will play and the scores are calculated by the automated referee.

Implementing a Cooperative Framework among Bio-inspired Robots based on Phonotaxis

José Camacho, Rosa M. de Molina, Eugenio Martín and Martín Mellado

This paper proposes the development of a cooperative framework among mobile robots, inspired in the phonotaxic behavior (tracking the source of a specific sound) observed in cricket mating. By means of this behavior, in combination with two other individual behaviors (for communication and obstacles avoidance) a set of five cooperation primitives is reached. A simulation platform has been used to test the design. Furthermore, two real robots, one acting as the female and the other one as the male, have been developed and tested: male emits a calling song (at a specific frequency) and female tracks or moves away from the sound source.

Auction like Task Allocation and Motion Coordination Strategies for Multi-Robot Transport Tasks

José Guerrero and Gabriel Oliver

In this paper we present a task allocation method based on auction mechanisms that allows to find how many robots are needed to execute a task. This number is unknown and depends on several factors. There are also different types of tasks that must be executed using different skills of the robots. It is very difficult to find a correct allocation under this conditions and at present it is an open problem. We also propose two motion coordination methods to reduce the interference effect between robots. To test our system a modification of the well know foraging task has been used. This task introduces special characteristics, not directly studied in previous work, that our method try to solve.

Rigidity and Persistence of Three and Higher Dimensional Formations

Julien M. Hendrickx, Baris Fidan, Changbin Yu, Brian D. O. Anderson and Vincent D. Blondel

In this paper, we generalize the notion of persistence, which has been originally introduced for two-dimensional formations, to R^d for d>=3, seeking to provide a theoretical framework for real world applications, which often are in three-dimensional space as opposed to the plane. We verify that many of the properties of rigid and/or persistent formations established in R^2 are also valid for higher dimensions. Analysing the closed subgraphs and directed paths in persistent graphs, we derive some further properties of persistent formations. We also provide an easily checkable necessary condition for persistence.

On Swarm Optimality In Dynamic And Symmetric Environments

Yaniv Altshuler, Israel A. Wagner and Alfred M. Bruckstein

The field of multi agents and multi robotics has become increasingly popular during the last two decades. The motivation behind multi agents based systems is that many tasks can be rather efficiently completed by using multiple simple autonomous agents (robots, software agents, etc.) instead of a single sophisticated one. Such systems are usually also more adaptive, scalable and robust than those based on a single, highly capable, unit. However, when examining such systems, one may be concerned of the price tag attached to the decentralized nature of swarm based approaches. Meaning, while we simplify designs and control mechanisms in order to save costs and computation resources, how far do our systems drift from optimality ? This work examines this issue by constructing an optimal algorithm for the Dynamic Cooperative Cleaners problem. The performance of the known SWEEP protocol is compared to this of the optimal algorithm. The results of this comparison show that as the problem gets harder, the performance of the SWEEP protocol gets closer to those of the optimal algorithm. The work also presents insightful results concerning optimal swarms in symmetric environments.

The Cooperative Hunters— Efficient Cooperative Search For Smart Targets Using UAV Swarms

Yaniv Altshuler, Vladimir Yanovsky, Israel A.Wagner and Alfred M. Bruckstein

This work examines the Cooperative Hunters problem, where a swarm of UAVs (unmanned aerial vehicles) is used for searching after one or more "smart targets" which are moving in a predefined area, while trying to avoid detection by the swarm. By arranging themselves into an efficient flight configuration, the UAVs optimizes their integrated sensing capability, and are thus capable of searching much larger territories than a group of uncooperative UAVs. This work presents two decentralized cooperative search algorithms which demonstrate major improvements over the algorithm and analysis presented in previous works. The first algorithm uses improved flying patterns which achieve superior search performance. An analytic optimality proof for the algorithm’s performance is presented. The second algorithm is a fault tolerant algorithm which allows the UAVs to search in areas whose shapes and sizes are unknown to the UAVs in advance (in comparison to previous works, which are designed for rectangular shapes whose dimensions are known to the swarm).

Robot Behavior Planning by a Coordinator Robot in a Symbiotic Autonomous Human-Robot System

Tao Zhang, MD Hasanuzzaman, Vuthichai Ampornaramveth and Haruki Ueno

This paper addresses robot behavior planning by a coordinator robot in a symbiotic autonomous human-robot system according to human request. The coordinator robot is constructed based on a knowledge model of coordination which is a description world to describe a physical domain world on coordination, including interpreter for the human request domain, coordination interface for the problem domain, coordination policy for the solution domain, etc. The coordination policy in the model is adopted for planning robot behaviors by the coordinator robot. By means of a software platform and distributed intelligent agents, the coordinator robot can be implemented and the robot behavior planning can be carried out according to human request. In this paper, a coordinator robot is actually constructed and it can autonomously plan robot behaviors in a symbiotic autonomous human-robot system. The experimental work demonstrates the effectiveness of the proposed method.

A Randomized Gathering Algorithm for Multiple Robots with Limited Sensing Capabilities

Noam Gordon, Israel A. Wagner and Alfred M. Bruckstein

Consider a swarm of weak, anonymous and homogeneous ant-robots (or a(ge)nts), lacking memory, orientation, and communication capabilities, and having myopic sensors that tell them the directions to nearby robots, but not the distance from them. We present a simple randomized algorithm which, when performed by all members of the swarm, gathers them in a small region. We explore the interesting global phenomena that occur during the process, as is evident from our simulations, and provide a proof for the proposed algorithm's correctness.

Predicting the State of Agents Using Motion and Occupancy Grids

David Ball and Gordon Wyeth

This paper presents an approach to predicting the future state of dynamic agents using an extension to occupancy grids that we have called motion and occupancy grids. These grids capture both the location and motion of an agent. The basis of the prediction is to model the behaviour of an agent as the probability of transitions between the states represented in the grid. Prediction can then be made using a Markov chain. The potential state explosion inherent in this method is overcome by dynamically building the state transition matrix in a tree structure. The results show that the system is able to rapidly develop probabilistic models of complex agents and predict their future state.

Maintenance and Drift of Attention in Human-Robot Communication

Jun Mukai and Michita Imai

We describe a human-robot communication system called ACS that includes spontaneous generation of robot's attention. Here, attention refers to an arbitrary policy for selecting behaviors. Although attention is necessary for robot to react to human utterances in human-robot communication, attention is usually defined a priori by the designers of human-robot communication systems, which prevents communication between humans and robots. The reason is that the reactions of such robots are fixed for specific situations so that humans are easy to predict the robots reactions. We therefore developed ACS to enable robots to generate their own attentions without predefined settings. We propose Feature Drift in ACS. With Feature Drift, the system has specific features of objects and dynamically maintains its attention based on this feature data. In particular, Feature Drift can change the attention spontaneously in over time, which solves the problem of fixed reaction. We implemented ACS in a communication robot, Robovie, and evaluated it. The results showed that the robot could maintain its own attention and react to human utterances according to this attention.

Hazard: A Framework Towards Connecting Artificial Intelligence and Robotics

Peter J. Andersson

The gaming industry has started to look for solutions in the Artificial intelligence (AI) research community and work has begun with common standards for integration. At the same time, few robotic systems in development use already developed AI frameworks and technologies. In this article, we present the development and evaluation of the Hazard framework that has been used to rapidly create simulations for development of cognitive systems. Implementations includes for example a dialogue system that transparently can connect to either an Unmanned Aerial Vehicle (UAV) or a simulated counterpart. Hazard is found suitable for developing simulations supporting high-level AI development and we identify and propose a solution to the factors that make the framework unsuitable for lower level robotic specific tasks such as event/chronicle recognition.

Cognitive Map Merging For Multi-Robot Navigation

Anderson A. Silva, Esther L. Colombini and Carlos H. C. Ribeiro

In this paper we investigate a map-building strategy based on geometric and topological information about the environment, acquired through sonar sensors and odometric actuator. To reduce robot individual exploration needs, a framework based on multi-robot map acquisition is proposed, where each robot executes a map bulding algorithm and perfoms exploration in the environment. The current global map is built based on the merging of overlapping regions among the previously cognitive maps. A brief description of the on-going research and the results obtained is also provided.

Mission Reliability Estimation for Repairable Robot Teams

Stephen B. Stancliff, John M. Dolan and Ashitey Trebi-Ollenu

NASA has expressed interest in using modular self-repairable robotic teams for the exploration and colonization of Mars. One of the reasons often given for using repairable robots is increased reliability. Analytical tools are needed for estimating the reliability of robotic missions in order to determine if this reasoning is correct, and for what types of missions. In this paper we present a method, based on standard tools from reliability engineering, for predicting the reliability of repairable robot teams. We then use this method to compare the reliability of repairable and nonrepairable robot teams for a simple mission scenario.

A Hybrid Evolutionary Probablistic Framework for Developing Robotic Team Behaviors

Edward Newett and Ashraf Saad

One of the inherent issues in team-based multiagent robotics is coordinating a cooperative task decomposition. Use of explicit communication models or game theoretic approaches to model teammate behaviors can be costly and error-prone. This paper describes a method of discovering a set of behaviors that allows a team to intrinsically function in a collaborative manner. Probabilistic planners based on spreading activation networks that determine these behaviors are implemented in each robot. A genetic algorithm is used to optimize the link strengths within each of these networks to produce an overall dynamic team. It is shown that a team controlled by spreading activation networks can perform well as a team by maintaining these behaviors in environmental situations other than the one used for GA evolution. From this framework, a goal-directed task planning approach can be envisioned to deploy a fully functional robot team.

Modifying Neuro Evolution For Mobile Robotic Behavior Development

Sekou Remy and Ashraf Saad

This work examines the effect of two modifications of typical approaches to evolving neuro controllers for robotic behaviors. First evolutionary methods were constrained to modify only one element of the population, the only element to be evaluated in the robot. Secondly the algorithm was allowed to incorporate genotypes provided by external sources. These modifications were evaluated through the use of a mobile robot simulator. Each was allowed to evolve in an arena that allowed it to interact with other robots. Experiments were conducted to investigate the effect of sharing genotypes and their corresponding fitness among homogeneous robots - the robots differed only in the initial random phenotype. The experiments showed that the ability to incorporate successful genotypes from others increased the rate at which evolution progressed. Communication of good genotypes allowed behaviors to get fitter faster.

Agent-Oriented Design of a Multi-Robot System

Bram Gruneir, Ben Miners, Alaa Khamis, Hamada Ghenniwa and Mohamed Kamel

This paper presents a new architecture for multiple robot systems using an agent oriented design methodology. The proposed architecture combines the hierarchical and the decentralized approaches. It splits all processes into two layers: the cognitive layer, where the higher brain functions take place, and the action layer, where the low level functions take place. It also addresses the use of cooperative software agents organized in hardware or software components. Each of these agents independently handles a small set of specialized tasks and cooperates to achieve system-level goals. The overall system behaviour emerges from the autonomous behaviours of the individual agents. This architecture defines clear boundaries between the processes that occur within a robot. Advantages include support for multiple robots with different specifications to communicate with each other and perform meaningful tasks. Experiments using mobile, autonomous robots equipped with a vision system are conducted to demonstrate the usefulness of the proposed architecture to facilitate the development of multi-robot cooperative behaviours.

Enabling Spoken Dialogue Interaction About Team Activities

Laura M. Hiatt and Lawrence Cavedon

Spoken language dialogue is a powerful mode for human-robot interaction in complex, dynamic environments. We describe extensions to an existing dialogue management system that enables activity-oriented interaction with multi-robot teams.

DIAMOND : A Physical Multiagent Systems Codesign Approach

Jean-Paul Jamont and Michel Occello

Multiagent systems are well suited to specify requirements for open physical complex systems. However, up to now, no method allows to build software/hardware hybrid multiagent systems. This paper presents an original method for designing physical multiagent systems. It advocates a basic multiagent phase able to tackle functionnal and organizational issues, associated to a componential phase for the detailed design making easier the software/hardware partitionment.

Structural Persistence of Three Dimensional Autonomous Formations

Changbin Yu, Julien M. Hendrickx, Baris Fidan and Brian D.O. Anderson

Built upon a recently developed theoretical framework, we consider some practical issues raised in multi-agent formation control in three dimensional space. We introduce the partial equilibrium problem, which is associated with unsafe or unstable control of a formation in practical 3-dimensional applications. We define structurally persistent graphs, a class of persistent graphs free of any partial equilibrium problem. In real deployment of control of multi-agent systems, formations with underlying structurally persistent graphs are of interest. We study the connections between the allocation of degrees of freedom (DOFs) across agents and the characteristics of persistence and/or structural persistence of a directed graph. We also show how to transfer degrees of freedom between agents, when the formation changes with new agent(s) added, to preserve persistence and/or structural persistence.