HYBRID SOM-SVM ALGORITHM FOR REAL TIME SERIES FORECASTING

Juan Manuel Górriz Sáez, Carlos García Puntonet and E. W. Lang

In this paper we show a new on-line parametric model for time series forecasting based on Vapnik-Chervonenkis (VC) theory. Us-ing the strong connection between support vector machines (SVM) and Regularization theory (RT), we propose a regularization operator in or-der to obtain a suitable expansion of radial basis functions (RBFs) with the corresponding expressions for updating neural parameters. This op-erator seeks for the “flattest” function in a feature space, minimizing the risk functional. Finally we mention some modifications and extensions that can be applied to control neural resources and select relevant input space.

AN ADAPTIVE SLIDING-MODE FUZZY CONTROL (ASMFC) APPROACH FOR A CLASS OF NONLINEAR SYSTEMS

Jian-Hua Zhang and Johann F. Böhme

This paper uses the concept of sliding-mode control (SMC), as a special approach in nonlinear control theory, in aiding the design of a fuzzy controller. The mathematical specifics of the presented approach are given along with its performance analysis. It was concluded that the new approach with distinctive characteristics holds potential for coping with difficult control problems for a class of complex (generally nonlinear) systems.

H CONTROL THEORY ON THE INFINITE DIMENSIONAL SPACE

Chuan-Gan Hu

In this paper, the VH^\infty control theory on an infinite dimensional algebra to itself is presented. In order to establish the VH^\infty control theory, the concept and the properties of a meromorphic mapping and the theory of VH^p spaces on an infinite dimensional algebra to itself are founded.

HBP: A NOVEL TECHNIQUE FOR DYNAMIC OPTIMIZATION OF THE FEED-FORWARD NEURAL NETWORK CONFIGURATION

Allan K.Y. Wong , Wilfred W.K. Lin and Tharam S. Dillon

The novel Hessian-based pruning (HBP) technique to optimize the feed-forward (FF) neural network (NN)configuration in a dynamic manner is proposed. It is then used to optimize the extant NNC (Neural Network Controller) as the verification exercise. The NNC is designed for dynamic buffer tuning to eliminate buffer overflow at the user/server level. The HBP optimization process is also dynamic and operates as a renewal process within the service life expectancy of the target FF neural network. Every optimization renewal cycle works with the original NN configuration. In the cycle all the insignificant NN connections are marked and then skipped in the subsequent operation before the next optimization cycle starts. The marking and skipping operations together characterize the dynamic pruning nature of the HBP. The interim optimized NN configuration produced by every HBP cycle is different, as the response to the current system dynamics. The verification results with the NNC indicate that the HBP technique is indeed effective because all the interim optimized/pruned NNC versions incessantly and consistently yield the same convergence precision to the original NNC predecessor, and with a shorter control cycle time.

DECENTRALIZED ESTIMATION FOR AGC OF POWER SYSTEMS

Xue-Bo Chen, Xiaohua Li and Srdjan S. Stankovic

A decentralized state estimation method for automatic generation control (AGC) of interconnected power systems is proposed in this paper. Based on the Inclusion Principle for linear stochastic systems, the state space model of the system is decomposed as a group of pair-wise subsystem models. The overlapping decentralized estimators and fully decentralized estimators are designed for each pair subsystems in the framework of LQG control schemes. Two types of estimators are considered for the cases of full and reduced measurement sets in the framework of system closed-loop operations. Simulation results show a high quality of the AGC scheme based on dynamic controllers with the proposed state estimators.

DEVICE INTEGRATION INTO AUTOMATION SYSTEMS WITH CONFIGURABLE DEVICE HANDLER

Anton Scheibelmasser, Udo Traussnigg, Georg Schindin and Ivo Derado

One of the most important topics in the field of automation systems is the integration of sensors, actuators,measurement devices and automation subsystems. Especially automation systems like test beds in the automotive industry impose high requirements regarding flexibility and reduced setup and integration time for new devices and operating modes. The core function of any automation system is the acquisition, evaluation and control of data received by sensors and sent to actuators. Sensors and actuators can be connected directly to the automation systems. In this case they are parameterised using specific software components, which determine the characteristics of every channel. In contrast to this, smart sensors, measurement devices or complex subsystems have to be integrated by means of different physical communication lines and protocols. The challenge for the automation system is to provide an integration platform, which will offer easy and flexible way for the integration of this type of devices. On the one hand, a sophisticated interface to the automation system should trigger, synchronise and evaluate values of different devices. On the other hand, a simple user interface for device integration should facilitate the flexible and straightforward device integration procedure for the customer. Configurable Device Handler is a software layer in the automation system, which offers a best trade-off between the complex functionality of intelligent devices and their integration in a simple, fast and flexible way. Due to a straightforward integration procedure, it is possible to integrate new devices and operation modes in a minimum of time by focusing on device functions and configuring the automation system,rather than writing software for specific device subsystems. This article gives an overview of Configurable Device Handler, which was implemented in a test bed automation system. It provides an insight into the architecture of the Configurable Device Handler and shows the principles and the ideas behind it. Finally,new aspects and future developments are discussed.

AN INSTRUMENT CONTROL SYSTEM USING PREDICTIVE MODELLING

Geoffrey Holmes and Dale Fletcher

We describe a system for providing early warning of possible error to an operator in control of an instrument providing results in batches from samples, for example, chemical elements found in soil or water samples. The system has the potential to be used with any form of instrument that provides multiple results for a given sample. The idea is to train models for each measurement, using historical data. The set of trained models are then capable of making predictions on new data based on the values of the other measurements. This approach has the potential to uncover previously unknown relationships between the measurements. An example application has been constructed that highlights the difference of the actual value for a measurement from its predicted value. The operator is provided with sliders to attenuate the sensitivity of the measurement perhaps based on its importance or its known sensitivity.

AN EVOLUTIONARY ALGORITHM FOR IDENTIFICATION OF NON-STATIONARY LINEAR PLANTS WITH TIME DELAY

Janusz P. Paplinski

The identification of time delay in the linear plant is one of the important tasks. It is especially hard problem when the plant is non-stationary. New possibility in this field is opened by application of an evolutionary algorithm. The method of identification proposed in the paper is based on three classes of input signals. In the first case we can obtain and operate on the whole unit step response. In the second way we operate on a random signal of control, and in the last we have the stairs input signal. The identification without and with disturbances is considered.

HIERARCHICAL MODAL CONTROL OF A NOVEL MANIPULATOR

Clarence de Silva and Jian Zhang

This paper focuses on the development and implementation of an intelligent hierarchical controller for the vibration control of a deployable manipulator. The emphasis is on the use of knowledge-based tuning of the low-level controller so as to improve the performance of the system. To this end, first a fuzzy inference system (FIS) is developed. The FIS is then combined with a conventional modal controller to construct a hierarchical control system. Specifically, a knowledge-based fuzzy system is used to tune the parameters of the modal controller. The effectiveness of the hierarchical control system is investigated through numerical simulation studies. Examples are considered where the system experiences vibrations due to initial disturbance at the flexible revolute joint or due to maneuvers of a deployable manipulator. The results show that the knowledge-based hierarchical control system is quite effective in suppressing vibrations induced due to the above mentioned disturbances. Results suggest that performance of the modal controller could be significantly improved through knowledge-based tuning.

AN EFFECTIVE APPROACH FOR REAL-WORLD PRODUCTION PLANNING

Jesuk Ko

This paper shows an application of constraint logic-based approach to the realistic scheduling problem. Operations scheduling, often influenced by diverse and conflicting constraints, is strongly NP-hard problem of combinatorial optimization. The problem is complicated further by real scheduling environments, where a variety of constraints in response are critical aspects for the application of a solution. Constraint logic programming technique well armed with the major function of constraint handling and solving mechanisms can be effectively applied to solve real-world scheduling problems. In this study, the scheduling problem addressed, based on a dye house involving jobs associated with the coloring of different fibers, is characterized by various constraints like color precedence, dye machine allocation and time constraints. The solution procedure used takes into account a number of dye house performance measures which include on-time delivery and resource utilization. The results indicate that constraint-based scheduling is computationally efficient in schedule generation in that a solution can be found within a few seconds. Furthermore, solutions produced always minimize the mean tardiness and maximize the utilization of dyeing facilities.

NON LINEAR SPECTRAL SDP METHOD FOR BMI-CONSTRAINED PROBLEMS: APPLICATIONS TO CONTROL DESIGN

Jean-Baptiste Thevenet, Dominikus Noll and Pierre Apkarian

The purpose of this paper is to examine a nonlinear spectral semidefinite programming method to solve problems with bilinear matrix inequality (BMI) constraints. Such optimization programs arise frequently in automatic control and are difficult to solve due to the inherent non-convexity. The method we discuss here is of augmented Lagrangian type and uses a succession of unconstrained subproblems to approximate the BMI optimization program. These tangent programs are solved by a trust region strategy. The method is tested against several difficult examples in feedback control synthesis.

RESISTANCE SPOT WELDING PROCESS IDENTIFICATION AND INITIALIZATION BASED ON SELF-ORGANIZING MAPS

Heli Junno, Perttu Laurinen, Eija Haapalainen, Lauri Tuovinen, Juha Röning, Dietmar Zettel, Daniel Sampaio, Norbert Link and Michael Peschl

Resistance spot welding is used to join two or more metal objects together, and the technique is in widespread use in, for example, the automotive and electrical industries. This paper discusses both the identification of different spot welding processes and the process initialization parameters leading to high-quality welding joints. The goal is to improve the quality of welding joints. In this research, self-organizing maps (SOMs) were used, and optimal features for the training parameters were sought. According to the results, processes can be classified by specific features. When introducing new data to trained SOMs, the welding operator can visually identify similar processes. After process identification, the most similar process is retrieved and a self-organizing map is trained for this specific process. The initialization parameters leading to successful welds in that process can thus be identified, which means that the manufacturers can use them to initialize their welding machines. It is concluded that self-organizing maps can be used to identify different spot welding processes and to find the appropriate initialization parameters for welding machines.

OPTIMIZATION OF CURRENT EXCITATION FOR PERMANENT MAGNET LINEAR SYNCHRONOUS MOTORS

Christof Röhrig

The main problem in improving the tracking performance of permanent magnet linear synchronous motors is the presence of force ripple caused by mismatched current excitation. This paper presents a method to optimize the current excitation of the motors in order to generate smooth force. The optimized phase current waveforms produce minimal ohmic losses and maximize motor efficiency. The current waveforms are valid for any velocity and any desired thrust force. The proposed optimization method consist of three stages. In every stage different harmonic waves of the force ripple are reduced. A comparison of the tracking performance with optimized waveforms and with sinusoidal waveforms shows the effectiveness of the method.

PARAMETER CONVERGENCE IN ADAPTIVE FUZZY CONTROL

Domenico Bellomo, David Naso and Robert Babuska

In this paper, the convergence of parameter estimates and the interactions among the two adaptive fuzzy systems constituting an indirect adaptive fuzzy controller are studied, both analytically and by means of simulations with a second-order nonlinear system. The analytical results and the simulations, performed with various initial conditions and learning rates, highlight how the interactions affect the behavior of the adaptive control scheme with regard to the control performance in terms of a tracking error and to the accuracy and relevance of the identified fuzzy models.

TABU SEARCH STRATEGIES IN SCHEDULING PROBLEM IN FLEXIBLE MANUFACTURING SYSTEM - Considering tool switches and number of setups

Antonio Gabriel Rodrigues, Arthur Tórgo Gómez

In this paper it’s investigated the impact of the Tabu List size, neighborhood generation approach and the managing of the decision variables of the Objective Function in the quality of a Tabu Search solution to the Scheduling Problem applied to a Flexible Manufacturing System. It was used a Part Scheduling Model, which starts with qualitatively different initial solutions that yields experiments in which it’s observed the Tabu List size influence in the results quality, according to the pre-defined Objective Function variables contribution. This Model creates a schedule in a Flexible Manufacturing System, considering resident tooling concepts, production turns, Part Selection, Machine Magazine Constraints and Due-dates. Numerical results show relations among neighborhood strategies and the Tabu List size behavior considering initial solutions and contribution managing of the Objective Function variables.

BEST-ACTION PLANNING FOR REAL-TIME RESPONSE - An approach in ORICA

Tariq Ali Omar, Ana Simonet, Michel Simonet

A planner for real-time response aims at building a plan to safely guide the world to its goal state by guaranteeing response deadlines. Ideally, it should find the best possible paths for the world. To achieve this ideal behaviour, it must be provided with maximum world behaviour characteristics and be able to control the response behaviour of the system under-control to its advantage. It should also be able to reason about one path with respect to the other, based on the execution duration, the amount of resources used and the system safety. In this paper, we present the ORICA (OSIRIS1 real-time intelligent control architecture) real-time response planner, which builds plans that permit the real-time system to strive to achieve its goal in un-guaranteed environment behaviour, while still ensuring system safety. It also possesses heuristic reasoning capability for comparison of different paths when a choice of path is possible.

A STOCHASTIC OFF LINE PLANNER OF OPTIMAL DYNAMIC MOTIONS FOR ROBOTIC MANIPULATORS

Taha Chettibi, Moussa Haddad, Samir Rebai and Abd Elfetah Hentout

we propose a general and simple method that handles free (or point-to-point) motion planning problem for redundant and non-redundant serial robots. The problem consists of linking two points in the operational space,under constraints on joint torques, jerks, accelerations, velocities and positions while minimizing a cost function involving significant physical parameters such as transfer time and joint torque quadratic average. The basic idea is to dissociate the search of optimal transfer time T from that of optimal motion parameters. Inherent constraints are then easily translated to bounds on the value of T. Furthermore, a stochastic optimization method is used which not only may find a better approximation of the global optimal motion than is usually obtained via traditional techniques but that also handles more complicated problems such as those involving discontinuous friction efforts and obstacle avoidance.

DISTRIBUTED LOAD BALANCING OF DISTRICT HEATING SYSTEMS - A SMALL-SCALE EXPERIMENT

Fredrik Wernstedt and Paul Davidsson

We present results from experiments where the effects of automatic flow control at a single substation is compared to automatic cooperative concurrent flow control at multiple substations. The latter approach is made possible by equipping individual substations with some computing power and integrating them into a communications network. Software agents, whose purpose is to cooperate with other software agents (substations) and to invoke reductions, are connected to each substation. The experiments show that it is possible to automatically load balance a small district heating network using agent technology, e.g., to perform automatic peak clipping and load shifting.

DYNAMIC BOOKING POLICY FOR AIRLINE SEAT INVENTORY CONTROL

Kristine Rozite, Nicholas A. Nechval, Konstantin N. Nechval and Edgars K. Vasermanis

It is common practice for airlines to charge several different fares for a common pool of seats. This paper presents the optimization algorithms that have been used to address the problem of when to refuse booking requests for a given fare level to save the seat for a potential request at a higher fare level. Dynamic and dynamic adaptive booking policies for multiple fare classes that share the same seating pool on one leg of an airline flight, when seats are booked in a nested fashion and when lower fare classes book before higher ones, are determined. The dynamic policy of airline booking makes repetitive use of a static method over the booking period, based on the most recent demand and capacity information. It allows one to allocate seats dynamically and anticipatory over time. The dynamic adaptive policy, in addition, deals with the case when only the functional forms of the probability density functions for reservation requests for various fare classes are given. In this case actual airline demand data are used to obtain estimates of expected demand for input into the airline optimization models, where we illustrate the practical importance of invariance for eliminating nuisance (unspecified) parameters from the problem. Although the traditional use of invariance has been in a decision theoretic setting, we instead use invariance to find a transformation of the data such that the distribution of the transformed data does not involve nuisance parameters. Illustrative examples are given.

FUZZY MODEL BASED CONTROL APPLIED TO IMAGE-BASED VISUAL SERVOING

Paulo Jorge Sequeira Gonçalves, Luís Mendonça, João Sousa and João Caldas Pinto

A new approach to eye-in-hand image-based visual servoing based on fuzzy modeling and control is proposed in this paper. Fuzzy modeling is applied to obtain an inverse model of the mapping between image features velocities and joints velocities, avoiding the necessity of inverting the Jacobian. An inverse model is identified for each trajectory using measurements data of a robotic manipulator, and it is directly used as a controller. As the inversion is not exact, steady-state errors must be compensated. This paper proposes the use of a fuzzy compensator to deal with this problem. The control scheme contains an inverse fuzzy model and a fuzzy compensator, which are applied to a robotic manipulator performing visual servoing, for a given profile of image features velocities. The obtained results show the effectiveness of the proposed control scheme: the fuzzy controller can follow a point-to-point pre-defined trajectory faster (or smoother) than the classic approach.

CONFLICT RESOLUTION FOR FREE FLIGHT CONSIDERING DEGREE OF DANGER AND CONCESSION

Mustafa Suphi Erden and Kemal Leblebicioğlu

In this study a conflict resolution technique based on danger and concession considerations is presented for free flight paradigm. A danger function which assigns a danger value for the conflict situation, and a concession function which assigns a concession value for the path followed by the aircraft are constructed. The danger and concession values are input to a fuzzy decision module. This module outputs the amount of deviation from the optimal path and the conflict is solved following these deviations. The method presented here is the third method we have been studying regarding to the conflict resolution problem. Its results are presented with a comparison to our other two studies.

QUALITATIVE AND QUANTITATIVE PROBABILISTIC TEMPORAL REASONING - for Industrial Applications

Gustavo Arroyo Figueroa

Many real-world domains, such industrial diagnosis, require an adequate representation that combines uncertainty and time. Research in this field involves the development of new knowledge representation and inference mechanisms to deal with uncertainty and time. However, current temporal probabilistic models become too complex when used for real world applications. In this paper, we propose a new model, Temporal Events Bayesian Networks (TEBN), based on a natural extension of a simple Bayesian network. TEBN tries to make a balance between expressiveness and computational efficiency. Based on a temporal node definition, causal-temporal dependencies are represented by qualitative and quantitative relations, using different time intervals within each variable (multiple granularity). Qualitative knowledge about temporal relations between variables is used to facilitate the acquisition of the quantitative parameters. The inference mechanism combines qualitative and quantitative reasoning. The proposed approach is applied to a thermal power plant through a detailed case study, with promising results.

GLOBAL CONDITION MONITORING SYSTEM - Implementing MATLAB-Based Analysis Services

Henri Helanterä, Mikko Salmenperä and Hannu Koivisto

Proactive maintenance is a solution to increase the availability of the production equipment in the process industry. It involves online condition monitoring of field devices and reliably diagnosing the reason behind any abnormal behaviour, thus helping to rationalise maintenance operations. If the huge amount of information available at the different industrial sites was available for analysis, significant improvements could be made to the predicting capabilities of condition monitoring and the accuracy of fault diagnostics. The global condition monitoring system architecture described in this paper is based on distributed agent-architecture and employs data communication networks to connect the industrial sites to one or more service centres. Many successful methods used in condition monitoring and fault diagnostics are based on various computational intelligence techniques and employing these methods often requires advanced tools. MATLAB software is a de facto standard in numerical computing but integrating MATLAB as a computing server to the J2EE-based condition monitoring system is a laborious task as no all-purpose and easy-to-use methods exist. However, this paper introduces some strategies to overcome the integration problem. The most important solution presented here is inverted calling scheme. Also two other approaches are discussed: using MATLAB engine functions via C-language native methods and deployment of stand-alone MATLAB COM components. All the above strategies have their advantages and weaknesses. Implementing the inverted call requires more effort from the programmer but is standard-compliant. Exploiting engine functions and COM components is easier as some ready-made software can be employed but the emerging solutions are not pure-Java.

AN EVOLUTIONARY APPROACH TO NONLINEAR DISCRETE - TIME OPTIMAL CONTROL WITH TERMINAL CONSTRAINTS

Yechiel Crispin

The nonlinear discrete-time optimal control problem with terminal constraints is treated using a new evolutionary approach which combines a genetic search for finding the control sequence with a solution of the initial value problem for the state variables. The main advantage of the method is that it does not require to obtain the solution of the adjoint problem which usually leads to a two-point boundary value problem combined with an optimality condition for finding the control sequence. The method is verified on two problems. The first problem is the discrete velocity direction programming with the effects of gravity and thrust, with a terminal constraint on the final vertical position. The second problem is an extension of the first problem to include the effect of viscous drag. The solutions of both problems compare favorably with the results of gradient methods.

A GROUP DECISION SUPPORT SYSTEM - A description on models and modules in GDSS based on cooperative MAS

Shi-Weiren, Jiang-Daoping, Liang-Yonglin and Chen-Jing

GDSS is popular and attractive topic in decision field. It is reasonable to combine multi-Agent technology and GDSS because they are both distributed systems and support interaction in group members. We propose models to describe character of Agent and issue in GDSS, define the workflow of group-decision as cognitive, group organizing, decision-making by cooperation, feedback and adjust decision, conduce consensus decision by negotiation, knowledge management and repository evolution, and explain the process of every part.

A DISTURBANCE COMPENSATION CONTROL FOR AN ACTIVE MAGNETIC BEARING SYSTEM BY A MULTIPLE FXLMS ALGORITHM

Min Sig Kang and Joon Lyou

In this paper, a design technique is proposed for a disturbance feedforward compensation control to attenuate disturbance responses in an active magnetic bearing system, which is subject to base motion. To eliminate the sensitivity of model accuracy to disturbance responses, the proposed design technique is an experimental feedforward compensator, developed from an adaptive estimation, by means of the Multiple Filtered-x least mean square (MFXLMS) algorithm. The compensation control is applied to a 2-DOF active magnetic bearing system subject to base motion. The feasibility of the proposed technique is illustrated, and the results of an experimental demonstration are shown.

AN INTELLIGENT RECOMMENDATION SYSTEM BASED ON FUZZY LOGIC

Shi Xiaowei

An intelligent recommendation system for a plurality of users based on fuzzy logic is presented. The architecture of a multi-agent recommendation system is described. How the system simulates human intelligence to provide recommendation to users is explained. The recommendation system is based on the fuzzy user profile, fuzzy filtering and recommendation agents. The user profile is updated dynamically based on the feedback information. Fuzzy logic is used in fuzzy filtering to integrate different types of features together for a better simulation of human intelligence. Ambiguity problems can be solved successfully in this system, e.g., deducing whether a programme with both interesting features and uninteresting features is worth recommending or not.

WHEELED VEHICLES CLASSIFICATION USING RADIAL BASE FUNCTION NEURAL NETWORK - Intelligent Control Systems and Optimization

Jerzy Jackowski and Roman Wantoch-Rekowski

The paper presents the problem of using neural network for military vehicle classification on the basis of ground vibration.

A REVIEW OF ADVANCES IN ECONOMIC DISPATCH USING ARTIFICIAL NEURAL NETWORKS

Tahir Nadeem Malik

Economic Dispatch Problem (EDP) has been discussed with reference to the developments based on mathematical programming techniques in general and Artificial Neural Networks (ANN) approaches in particular. Brief survey has been included on the Economic Dispatch in mathematical programming and optimization techniques domain. A selected survey / overview on Economic Dispatch using Artificial Neural Network within the IEE/IEEE publications frame work have been presented.

A TORQUE ESTIMATION METHOD TO AID AN INTELLIGENT MANAGEMENT SYSTEM FOR OIL WELLS AUTOMATED

Alberto S. Rebouças, Flávia N. Serafim, Milena de A. Moreira, Venício R. V. Rodeiro, Amauri Oliveira and Jés J. F. Cerqueira

This article presents a contribution for an intelligent management system for oil wells called SGPA that nowadays manages about 700 oil wells using the rod pumping lift method at Bahia State, Brazil. The intelligent management system will be applied on oil wells using the gradual pumping method. In this type of oil pumping method, the torque on the rod is very importance for detection of operational problems. It will be considered that the well is driven by an induction motor. A torque estimation method on rod and some results from laboratory are presented.

AN HORIZONTAL APPROACH TO BATCH SCHEDULING - Using the Simultaneous Manufacturing philosophy

Ana Almeida, Carlos Ramos and Sílvio do Carmo Silva

This paper is concerned with Batch Scheduling in job-shop like manufacturing systems. The Horizontal Scheduling approach is used, assuming that full scheduling of a simple or complex job, based on the job routing network of operations, from the first operation to the last, is performed before another job is considered for scheduling, having in consideration existing manufacturing processors and their availability. We follow this approach because we aim at compressing job flow time to a minimum as a strategy to meeting job due dates. To further enhance this objective the idea behind Simultaneous Manufacturing through, the widespread use of batch overlapping with Job Scheduling Patterns, which proved particularly effective in reducing job throughput time, maintaining operating simplicity and requiring reduced coordination

SCALED GRADIENT DESCENT LEARNING RATE - Reinforcement learning with light-seeking robot

Kary Främling

Adaptive behaviour through machine learning is challenging in many real-world applications such as robotics. This is because learning has to be rapid enough to be performed in real time and to avoid damage to the robot. Models using linear function approximation are interesting in such tasks because they offer rapid learning and have small memory and processing requirements. Adalines are a simple model for gradient descent learning with linear function approximation. However, the performance of gradient descent learning even with a linear model greatly depends on identifying a good value for the learning rate to use. In this paper it is shown that the learning rate should be scaled as a function of the current input values. A scaled learning rate makes it possible to avoid weight oscillations without slowing down learning. The advantages of using the scaled learning rate are illustrated using a robot that learns to navigate towards a light source. This light-seeking robot performs a Reinforcement Learning task, where the robot collects training samples by exploring the environment, i.e. taking actions and learning from their result by a trial-and-error procedure.

SOLVING THE LONGEST WORD-CHAIN PROBLEM

Nobuo Inui, Yuji Shinano, Yuusuke Kounoike and Yosiyuki Kotani

This paper describes the definition of the longest SIRITORI problem as a problem of graph and the solution based on the integer problem (IP). This formulation requires large numbers of variables in proportion to the exponential order. Against this issue, we propose a solution based on the LP-based branch-and-bound method, which gradually solves the relaxation problems. This method is able to calculate the longest SIRITORI sequences for 130 thousand words dictionary within a second. In this paper, we compare the performances for the local-heuristic search and investigate the results for several conditions to explore the longest SIRITORI problem.

MODELLING THE MAN MACHINE INTERACTION - Erotetic Logic and Information Retrieval Systems

Antonio Bellacicco and Mario Vacca

The usual communication between man and machine is a one way interaction. It can be upgraded considering a two way interaction if the basic constituents of an information retrieval system are deeply modified in their principles. In this paper we redefine, in a syntactical way, some concepts of the erotetic logic to make them more easily computable and show how them can be used to solve some problems in the field of information retrieval systems. The result is the possibility to build more flexible and powerful information retrieval systems.

MULTI-AGENTS BASED REFERENCE MODEL FOR FAULT MANAGEMENT SYSTEM IN INDUSTRIAL PROCESSES

Mariela Cerrada-Lozada, Juan Cardillo, Jose Aguilar-Castro and Raúl Faneite

Nowadays, industrial necessities claims global management procedures integrating information systems in order to manage and to use the controlled-processes information and thus, to assure a good process behaviour. These aspects aim to the development of fault detection and diagnosis systems and making-decision systems. In this work, a reference model for fault management in industrial processes is proposed. This model is based on a generic framework using multi-agent systems for distributed control systems; in this sense, the fault management problem is viewed like a feedback control process and the actions are related to the making-decision in the preventive maintenance task scheduling and the running of preventive and corrective specific maintenance tasks. A particular methodology permitting the conception and analysis of the agent systems is used for the agents design. As a result, a set of models describing the general characteristics of the agents, specific tasks, communications and coordination is obtained.

FUZZY CONTROL OF FABRICS DRYING ON AN INDUCTION HEATED ROTATING CYLINDER: Experimental results

Sergio Pérez, Zulay Niño, Normand Thérien and Arthur D. Broadbent

The removal of water from materials in textile industry and pulp and paper industry requires a high-energy consumption, increasing significantly the operating costs. Nevertheless, electromagnetic induction heating is an alternative with considerable potential for the thermal treatment of materials. Specifically, heating the surface of a metallic cylinder by electromagnetic induction has opened up a range of applications for continuos heating, pre-drying and drying of fibrous web. Otherwise, these news electrotechnologies with industrial applications have to be used under controlled operational conditions. The past few years witnessed a rapid growth in the use of fuzzy logic controllers for the control of processes, which are complex and ill defined. These control systems are inspired by the approximate reasoning capabilities of the process operator. The purpose of this paper is to improve and apply an digital control structure on the basis of fuzzy logic technique for the textile drying using a rotational cylinder heated by electromagnetic induction, manipulating the power supply to the inductors to control the exit humidity of the web. The proposed fuzzy logic controller was tested experimentally in a dryer pilot-scale plant and the results show the capability of the controller to reach the set point initially fixed at 20 g water/100 g dry fabric. Once reached the set point, continuing the trial, steps changes of the web-cylinder contact surface and the set point were done and the results shows the stability of the proposed fuzzy logic controller in both perturbations.

GENETIC ALGORITHMS APPLIED TO THE OPTIMIZATION OF GASIFICATION FOR A GIVEN FUEL

Miguel Caldas, Luisa G. Caldas and Viriato Semião

Gasification is a well-known technology that allows for a combustible gas to be obtained from a carbonaceous fuel by a partial oxidation process (POX). The resulting gas (synthesis gas or syngas) can be used either as a fuel or as a feedstock for chemical production. Recently, gasification has also received a great deal of attention concerning power production possibilities through IGCC process (Integrated Gasification Combined Cycle), which is currently the most environmentally friendly and efficient method for the production of electricity. Gasification allows for low grade fuels, or dirty fuels, to be used in an environmental acceptable way. Amongst these fuels are wastes from the petrochemical and other industries, which vary in composition shipment to shipment and from lot to lot. If operating conditions are kept constant this could result in lose of efficiency. This paper presents an application of Genetic Algorithms to optimise the operating parameters of a gasifier processing a given fuel. Two different objective functions are used: one to be used if hydrogen production if the main goal of gasification; other to be used when power/heat production is the process’s aim. The optimization method developed could be used for on-line adjustment of the gasification operating parameters for each fuel lot, or shipment, thus improving overall performance of the industrial process.

MODEL REFERENCE CONTROL IN INVENTORY AND SUPPLY CHAIN MANAGEMENT - The implementation of a more suitable cost function

Heikki Rasku, Juuso Rantala and Hannu Koivisto

A method of model reference control is investigated in this study in order to present a more suitable method of controlling an inventory or a supply chain. The problem of difficult determining of the cost of change made in the control in supply chain related systems is studied and a solution presented. Both model predictive controller and a model reference controller are implemented in order to simulate results. Advantages of model reference control in supply chain related control are presented. Also a new way of implementing supply chain simulators is presented and used in the simulations.

FIPA-OS AGENTS APPLIED TO PROCESS SCHEDULING IN REALTIME MONITORING

Angel Gómez, Diego Cantorna, Carlos Dafonte and Bernardino Arcay

This work presents a mechanism for the management of network tasks, based on the technology of Intelligent Agents applied to a project of Telemedicine in Intensive Care Units (ICUs).The telemedicine system provides the real time acquisition and analysis of physiological data of patients, the graphical visualisation of these data and their transmission to a central system charged with the collection and control of all the information concerning the patient, including knowledge based systems for medical reasoning. The system tasks are managed through the use of intelligent agents, implemented according to the FIPA standard. Each of the agents disposes of a knowledge-based system for its decision-making.

AGENTS COORDINATION IN FLAT HIERARCHICAL SOCIETY-ORIENTED SYSTEMS

Mihaela R. Cistelecan

The paper aims to propose a framework that make possible to engineer a coherent society-oriented system. For this purpose the paper investigates the opportunity of importing the concept of sliding-mode control from systems theory into the society-oriented systems. The agent is modeled as a polynomial system. Differential algebra is used as an aggregation tool for different concepts.

DATA SECURITY CONSIDERATIONS IN MODERN AUTOMATION NETWORKS

Mikko Salmenperä and Jari Seppälä

The automation manufacturing business has reached its turning point and manufacturers are forced to create new business areas. Their expertise about field devices will be the source for future growth of automation industry. This includes monitoring, maintenance, data analysis and process tuning which all require good remoting capabilities in order to be successfully and cost efficiently applied as a service for production plants. This trend builds new challenges for automation services support systems. They are forced to adapt into global business model where customers utilise network connections from old modem lines into modern mobile communication networks. "Information is power" therefore securing production and other process related information systems in modern automation networks is becoming a necessity. The resent headlines have proved that automation systems are becoming more vulnerable with the inclusion of standard office and Internet technologies into the automation networks. The only way to meet the security challenges in a global scale is to build the the connectivity using secure-by-desing methodology.

OPTIMAL DESIGN OF VARIABLE STRUCTURE LOAD FREQUENCY CONTROLLER WITH NONLINEARITIES USING TABU SEARCH ALGORITHM

Naji A. Al-Musabi, Hussain N. Al-Duwaish, A. Mantawy, Zakariya Al-Hamouz and Samir Al-Baiyat

Optimal design of Variable Structure Controller (VSC) applied to Load Frequency Control (LFC) is explored in this paper. The controller was designed by an optimal method utilizing Tabu Search (TS) algorithm. The proposed method has been applied to a single nonreheat LFC area. Nonlinearities in the form of Generation Rate Constraint (GRC) and governor deadband backlash were included in the LFC model. Conventional optimal design methods are not efficient in designing controllers for models with nonlinearities. The new optimal approach using Tabu search algorithm has been applied efficiently to these models and compared to other methods reported in literature. Simulation results show that an improved robust dynamic behaviour can be achieved with the new optimal design method.

HOW TO ESCAPE TRAPS USING CLONAL SELECTION ALGORITHMS

Vincenzo Cutello, Giuseppe Narzisi, Giuseppe Nicosia, Mario Pavone and Giuseppe Sorace

The paper presents an experimental study on clonal selection algorithms (CSAs) to optimize simple and complex trap functions. Have been tested several setting of the proposed immune algorithms to effectively face this no easy computational problem. The key feature to solve the trap functions, hence escape traps, is the usage of the hypermacromutation operator couple with a traditional perturbation immune operator, the inversely proportional to the fitness function values hypermutation operator or the static hypermutation. The experimental results show that the CSA we designed is very competitive with the best algorithm in literature.

TUNING THE PARAMETERS OF A CLASSIFIER FOR FAULT DIAGNOSIS - Particle Swarm Optimization vs Genetic Algorithms

Cosmin Danut Bocaniala and José Sa da Costa

This paper presents a comparison between the use of particle swarm optimization and the use of genetic algorithms for tuning the parameters of a novel fuzzy classifier. In previous work on the classifier, the large amount of time needed by genetic algorithms has been significantly diminished by using an optimized initial population. Even with this improvement, the time spent on tuning the parameters is still very large. The present comparison suggests that using particle swarm optimization may improve considerably the time needed for tuning the parameters. In this way, the fuzzy classifier becomes suitable for real world application. The result is validated by application to a fault diagnosis benchmark.

ITERATIVE LINEAR QUADRATIC REGULATOR DESIGN FOR NONLINEAR BIOLOGICAL MOVEMENT SYSTEMS

Weiwei Li and Emanuel Todorov

This paper presents an Iterative Linear Quadratic Regulator (ILQR) method for locally-optimal feedback control of nonlinear dynamical systems. The method is applied to a musculo-skeletal arm model with 10 state dimensions and 6 controls, and is used to compute energy-optimal reaching movements. Numerical comparisons with three existing methods demonstrate that the new method converges substantially faster and finds slightly better solutions.

A COMBINED APPROACH TO FAULT DIAGNOSIS IN DYNAMIC SYSTEMS - Application to the Three-Tank Benchmark

Luís Palma, Fernando Coito and Rui Silva

This paper presents a combined approach to fault diagnosis (FDI) in discrete-time dynamic systems. The approach integrates classical and soft computing techniques for FDI. The typical methods based on signal models, and process models for residual generation are considered: parity equations, observers and parameter estimation. The role of integration of classical and intelligent techniques is enhanced. The proposed approach is applied to a typical nonlinear feed-water system – the three-tank benchmark. The three typical fault scenarios (actuator and component faults) defined in the benchmark problem are considered in this work.

DYNAMIC ROUTING AND QUEUE MANAGEMENT VIA BUNDLE SUBGRADIENT METHODS

Almir Mutapcic, Majid Emami and Keyvan Mohajer

In this paper we propose a completely distributed dynamic network routing algorithm that simultaneously regulates queue sizes across the network. The algorithm is distributed since each node decides on its outgoing link flows based only on its own and its immediate neighbors' information. Therefore, this routing method will be adaptive and robust to changes in network topology, such as the node or link failures. This algorithm is based on the idea of bundle subgradient methods, which accelerate convergence when applied to regular non-differentiable optimization problems. In the optimal network flow framework, we show that queues can be treated as subgradient accumulations and thus bundle subgradient methods also drive average queue sizes to zero. We prove the convergence of our proposed algorithm and we state stability conditions for constant step size update rules. Algorithm is implemented using Matlab and its performance is analyzed on a test network with varying data traffic patterns.

EVOLUTIONARY APPROACH FOR DYNAMIC SCHEDULING IN MANUFACTURING

Ana Maria Madureira, Carlos Ramos and Sílvio do Carmo Silva

This paper presents a simple and general framework exploring the potential of evolutionary algorithms, which is of practical utility, embedded in a simple framework to solve difficult problems in dynamic environments. The proposed evolutionary approach is in line with reality and away from the approaches that deal with static and classic or basic Job-Shop scheduling problems. In fact, in real world, where problems are essentially of dynamic and stochastic nature, the traditional methods or algorithms are of very little use. This is the case with most algorithms for solving the so-called static scheduling problem for different setting of both single and multi-machine systems arrangements. This reality, motivated us to concentrate on tools, which could deal with such dynamic, disturbed scheduling problems, both for single and multi-machine manufacturing settings, even though, due to the complexity of these problems, optimal solutions may not be possible to find. We decided to address the problem drawing upon the potential of Genetic Algorithms to deal with such complex situations. The paper describes a scheduling system, based on Genetic Algorithms to solve the Extended Job-Shop Scheduling Problem, where the products (jobs) to be processed have due dates, release times, different assembly levels and where random perturbations may occur over time.

AN LMI OPTIMIZATION APPROACH FOR GUARANTEED COST CONTROL OF SYSTEMS WITH STATE AND INPUT DELAYS

Olga I. Kosmidou, Y. S. Boutalis and Ch. Hatzis

The robust control problem for linear systems with parameter uncertainties and time-varying delays is examined. By using an appropriate uncertainty description, a linear state feedback control law is found ensuring the closed-loop system's stability and a performance measure, in terms of the {\it guaranteed cost}. An LMI objective minimization approach allows to determine the "optimal" choice of free parameters in the uncertainty description, leading to the minimal guaranteed cost.

AN LMI-BASED GENETIC ALGORITHM FOR GUARANTEED COST CONTROL

George A. Papakostas, Olga I. Kosmidou and I. E. Antonakis

In this paper a new approach for the Guaranteed Cost Control Problem (GCCP) is presented, using two efficient tools, Linear Matrix Inequalities (LMIs) and Genetic Algorithms (GAs). A linear system with parametric uncertainty is considered for which a control law is to be found, minimizing a performance index. In a previous paper, an efficient method has been presented by using an LMI optimisation technique. A combined use of LMIs and GAs is proposed in the present approach that allows further improvement of the design procedure.

USING A DISCRETE-EVENT SYSTEM FORMALISM FOR THE MULTI-AGENT CONTROL OF MANUFACTURING SYSTEMS

Guido Maione and David Naso

In the area of Heterarchical Manufacturing Systems Modeling and Control, a relatively new paradigm is that of Multi-Agent Systems. Many efforts have been made to define the autonomous agents concurrently operating in the system and the relations between them. But few results in the current literature define a formal and unambiguous way to model a Multi-Agent System, which can be used for the real-time simulation and control of flexible manufacturing environments. To this aim, this paper extends and develops some results previously obtained by the same authors, to define a discrete event system model of the main distributed agents controlling a manufacturing system. The main mechanism of interaction between three classes of agents is presented.

FROM UML TOWARDS PETRI NETS TO SPECIFY AND VERIFY

Thouraya Bouabana-Tebibel and Mounira Belmesk

UML nowadays, has emerged as the industry standard for object-oriented modeling. However, it still lacks s a well-defined semantic base enabling it to perform formal verification and validation tasks. Our goal being to provide system designers a life cycle of software development integrating conviviality and rigor, we propose a methodology to specify, verify and validate using UML. This methodology is based on a technique which derives colored Petri nets from UML class, statechart and collaboration diagrams. The approach that we propose allows the association of the formalization of the object dynamics with the formalization of the object behavior . A case study is provided to illustrate this technique.

REAL-TIME POSITION CONTROL OF A PNEUMATIC SYSTEM USING MODEL PREDICTIVE CONTROL

Doruk Akyıldız, Umut Karahan and Can Özsoy

Studies on the precise control applications with pneumatic systems have been growing in recent years.In addition to this, due to the complexity and non-linearity of the system the expected performance will only be gained by applying modern control strategies. So the subject of this paper is identification and real-time model predictive control of a pneumatic system. In order to realise system identification, a white noise signal is sent to the plant and the displacement outputs are stored. Afterwards these data are digitally processed and the parametric single-input single-output step response model is obtained. In the previous study on this system with a PD controller, a steady-state error is observed. In order to eradicate this, a Model Predictive Control – Dynamic Matrix Control algorithm is applied. To run this, in real-time, a programme is written in Matlab - Simulink and by using the code generated by Matlab - Real-Time Workshop, the real-time position control of the system is performed.

A DECENTRALIZED ROUTE GUIDANCE ALGORITHM IN URBAN TRANSPORTATION NETWORKS

Ludovica Adacher and Gaia Nicosia

In the last decades, due to the increasing car traffic and the limited capacity of urban networks, algorithms for the traffic management and route guidance are becoming more and more important. GPS technology can be used for fleet monitoring in urban or suburban areas, from a central monitoring station (reference station) and may provide useful information concerning the movement of all vehicles. Current route guidance systems are simple from an algorithmic point of view (they compute shortest paths to the destination), but they have to deal with huge size networks. For this reason, a decentralized approach, in which each vehicle independently calculates its own route, is desirable. Naturally, to limit the congestion due the single vehicles decisions, an estimate on the different possible routes is required. Hence, we propose a decentralized algorithm in which each vehicle computes its own route on the basis of the traffic information provided by the reference station. In order to cause the vehicle to choose different paths, the algorithm is allowed to take random decisions and the reference station is informed on the routes of all vehicles, so that traffic forecast is updated.

INTELLIGENT ELECTRIC DRIVE SYSTEM - A mechatronics approach

Guy Reginald Dunlop

Drive efficiency is an important consideration in most robotic applications. A hybrid controller for permanent magnet DC motors has been developed to control the current, and hence the output torque of the motor. An H bridge is used to provide the basic PWM voltage to the motor, and the controller switches the bridge between bipolar and unipolar modes in order to minimise the switching losses within the bridge and motor, and also to minimise the electromagnetic interference. The first application presented is for a walking robot, and the second is for a dexterous robotic hand. In both cases, control is obtained from a voltage sourced inverter by means of a tight control loop that uses position readings to infer velocity so that a full DC motor model can be utilised in the control computer. For the dexterous hand, current control is by model prediction to avoid the need for direct measurement. The controller and communications are contained within a small programmable system on a chip which together with a dual H bridge driver is integrated into small circuit board that is used for distributed control within the hand.

WAVE-WP (WORLD PROCESSING) TECHNOLOGY

Peter Sapaty and Masanori Sugisaka

A new computational and control model of parallel and distributed nature has been developed. It comprises self-evolving, space-conquering automaton, high-level system navigation and coordination language, describing system problems in a spatial pattern-matching mode, and related distributed control mechanisms for management of physical, virtual, and combined worlds. The model allows us to obtain complex spatial solutions in a compact, integral, and seamless way. It can be effectively used for the creation, integration, simulation, processing, management and control of a variety of dynamic and open systems – from physical to biological, and from artificial to natural.

AIRCRAFT ELECTRICAL MONITORING & FAULT DETECTION

George Kusic

Fault detection and electrical system monitoring (management) for aircraft/spacecraft dc or 400 Hz electrical systems is presented. Real-time ‘snapshot’ data is collected from current and voltage measurement transducers on radial or loop aircraft electrical system and introduced into a State Estimator. The State Estimator ‘smoothes’ the data, detects bad transducers, and calculates the best estimate of the voltage and phase angle at busses of the network, i.e., the ‘state’ of the network. Experimental results of estimation and fault detection are presented.

CONTROL THROUGH STATE CONVERGENCE OF TELEOPERATION SYSTEMS WITH VARYING TIME DELAY

José M. Azorín, Oscar Reinoso, José M. Sabater and Rafael Aracil

Teleoperation systems that use Internet as communication channel must deal with varying time delays. In these situations, the system can become unstable due to the irregular variations of the time delay. In this paper, a control method of teleoperation systems that we presented considering constant time delays is applied to control a teleoperation system with varying time delays. The control gains obtained with a constant time delay can be used to control the teleoperation system with varying time delays because of the control method robustness. Simulation and experimental results are presented to illustrate the validity of the method.

A SHALLOW DRAFT VEHICLE FOR INTERDISCIPLINARY RESEARCH AND EDUCATION

Carl Steidley, Ray Bachnak, Wien Lohatchit, Alex Sadovsky, and Cody Ross

Water quality data collection in shallow water areas can be a challenging task. Obstacles encountered in such environments include difficulty in covering large territories and the presence of inaccessible areas due to a variety of reasons such as a soft bottom or contamination. There is also a high probability of disturbing the test area while placing the sensors. This paper describes a NASA-funded project, which has had a great deal of student involvement and is currently in the test phase, to develop a remote-controlled, shallow-draft vehicle designed as a supplemental tool for our studies of the South Texas Coastal waters. The system transmits environmental data wirelessly via a radio to a docking and control station in real-time.

ANALYSIS OF THE ARCHITECTURE AND RELIABILITY OF DATA TRANSMISSION NETWORK USED FOR RADIO BASED CAB SIGNALING SYSTEM

Wang Junfeng, Zhang Yong, Wang Huashen and Wang Xishi

The application background and basic structure of train control system based on the combination of Radio Based Cab Signaling (RBCS) and Automatic Train Protection (ATP) is introduced. The architecture of the data transmission network used for RBCS is analyzed in detail, together with the reliability of radio data transmission.

MOTION PLANNING APPROACH OF A MULTI-FINGERED ROBOT FOR CARTON FOLDING OPERATIONS

Hidetsugu Terada and Takayuki Kobayashi

The motion planning approach of a multi-fingered robot for carton folding operations has been newly developed. This approach considers the loci of the tool center point for carton flap folding operations. Also that considers the push or fixing points of the carton flap. This approach is calculated from the rotating angle of a carton flap folding and the contact position with robot finger head and carton surface, using inverse kinematics. And this approach can be applied to changes of a carton size or a folding position. In cases in which the carton flap is folded using this approach, the robot finger head touches the carton surface without slipping and moves along circular continuous path. Therefore in case of the rectangular carton box folding, each robot finger moves in each 2.5-dimensional Cartesian frame. In this report, the proposed approach is verified using a prototype robot system. This prototype system consists of two pairs of the robot fingers and rotating mechanism for carton paper. Each finger has a 3-DOF SCARA type robot and a 1-DOF linear motion system. The testing carton boxes can fold to the desired shape.

TWO APPROACHES FOR A SERVOMECHANISM CONTROL SYSTEM USING COMPUTER VISION

João Manual R. S. Tavares, Ricardo Ferreira and Francisco Freitas

In this paper a servomechanism control system based on computational vision is presented. The control Images, based in hand language, are acquired by a generic webcam and processed, in the working phase, in quasi real time. For this processing, two approaches were considered: in the first one, we used the object control moments to identify the desired order; in the second approach, we used Images control orientation histograms. In both approaches, the preset orders Images to be considered for the servomechanism vision control are acquired in the learning phase. The used servomechanism and the two approaches used for the vision control system are described and some vantages and weakness of each are indicated. An example of an image control order set, which works satisfactory, is also presented and some conclusions and future works are also addressed.

ON THE DECENTRALIZED CONTROL OF LARGE DYNAMICAL COMPLEX SYSTEM

M. Kidouche, M. Zelmat and A. Charef

This paper describes a systematic procedure to build reduced order analytical models for a design of decentralized controllers for large scale interconnected dynamical systems. The design method employs Davison techniques to affect decoupling of the interconnections into its subsystems components which is done by using the most dominant eigenvalues and the most influent inputs in each subsystem. In this way, advantage can be taken of the special structural feature of a given system to devise feasible and efficient decentralized strategies for solving large control problem which are impractical to solve by one shot centralized methods.