CCI
CCI
CCI
COMPUTATIONAL
INTELLIGENCE
Proceedings of the EUROPEAN COMPUTING CONFERENCE
(ECC '09)
Proceedings of the 3rd International Conference on
COMPUTATIONAL INTELLIGENCE (CI '09)
Tbilisi, Georgia
June 26-28, 2009
Tbilisi, Georgia
June 26-28, 2009
Electrical and Computer Engineering Series
A Series of Reference Books and Textbooks
All the copyright of the present book belongs to the World Scientific and Engineering Academy and
Society Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or
otherwise, without the prior written permission of the Editor of World Scientific and Engineering Academy
and Society Press.
All papers of the present volume were peer reviewed by two independent reviewers. Acceptance was
granted when both reviewers' recommendations were positive.
See also: http://www.worldses.org/review/index.html
ISSN: 1790-5117
ISBN: 978-960-474-088-8
Tbilisi, Georgia
June 26-28, 2009
Editors:
Prof. Gia Sirbiladze, Iv. Javakhishvili Tbilisi State University, Georgia
Prof. Anna Sikharulidze, Iv. Javakhishvili Tbilisi State University, Georgia
Prof. Irina Khutsishvili, Iv. Javakhishvili Tbilisi State University, Georgia
Prof. Tamar Lominadze, Georgian Technical University, Georgia
Its success is reflected in the papers received, with participants coming from several countries,
allowing a real multinational multicultural exchange of experiences and ideas.
The accepted papers of this conference are published in this Book that will be indexed by ISI.
Please, check it: www.worldses.org/indexes as well as in the CD-ROM Proceedings. They will
be also available in the E-Library of the WSEAS. The best papers will be also promoted in many
Journals for further evaluation.
A Conference such as this can only succeed as a team effort, so the Editors want to thank the
International Scientific Committee and the Reviewers for their excellent work in reviewing the
papers as well as their invaluable input and advice.
The Editors
COMPUTING and COMPUTATIONAL INTELLIGENCE
Table of Contents
Plenary Lecture 1: Evolutionary Computation using Logic Programming 16
Jorge Ribeiro
Plenary Lecture 7: A Fuzzy Identification Problem for the Stationary Discrete Extremal Fuzzy 22
Dynamic System
Gia Sirbiladze
Plenary Lecture 8: On the Optimality of a General Production Lot Size Inventory Models with 23
Variable Parameters
Zaid Balkhi
A New Method for Controlling Boiler of Thermal Power Plant using Fuzzy Logic 31
Hamid Bentarzi, Rabah Amr Chentir, Nikos E. Mastorakis
Quality of the Information: The Application in the Winification Process in Wine Production 62
Jorge Ribeiro, Paulo Novais, Jose Neves, Manuel Delgado
Neural Network Approximating Nonlinear Sand Line Change from Aerial Survey Data Sets 91
T. Kerh, Y. M. Wang, G. S. Hsu, D. Gunaratnam
Fast Time Delay Neural Networks for Word Detection in Video Conference 120
Hazem M. El-Bakry, Nikos Mastorakis
On the Optimality of a General Production Lot Size Inventory Model with Variable Parameters 145
Zaid T. Balkhi
Artificial Neural Networks applied to Sensorless Control in a Switched Reluctance Motor 152
Eleonora Darie, Costin Cepisca, Emanuel Darie
Experimental Analysis of Advanced System for Reducing the Energy Consumption of Public 160
Street Lighting Systems
Costin Cepisca, Horia Andrei, Dogaru Valentin Ulieru, Laurentiu Stancu, Eleonora Darie
Fluid Approximation to Controlled Jump Markov Processes with Local Transitions 175
Alexey Piunovskiy
Mobile Platform for Testing Electric Traction Motor Prototypes in Real Road Conditions 187
Stefan Gheorghe, Mircea Covrig, Grigore Danciu, Costin Cepisca, Tudor Ursu, Soriun Dan
Grigorescu, Sanda Victorinne Paturca, Nicolae Jula, Dorin Oprea, Daniel Serban
Artificial Neural Networks applied to Measurements Rotor Angle in a Synchronous Generator 193
Eleonora Darie, Costin Cepisca, Emanuel Darie
Contributions Concerning the Sensitivity Analysis of the Analogue Two-Ports Circuits for 197
Communications
Horia Andrei, Costin Cepisca, Nicolae Jula, Paul Cristian Andrei, Laurentiu Stancu
Fuzzy Logic System for Variable Message Signs in Kuala Lumpur 209
Arash Moradkhani Roshandeh, Majid Joshani, Othman Che Puan
Improving Height-Balance in Search Trees: Center versus Root, Radius Versus Height 256
Koba Gelashvili, Irina Khutsishvili
Transition from MSN (Multi Service Network) Conception to NGN (Next Generation Networks) 290
Networks Conception
Jemal Beridze, Tatiana Burkadze
Development and Research of the Computer System Models Supporting the Human Resource 294
Selection for the Project Management
Gia Surguladze, Ekaterine Turkia, Nino Topuria, Megi Giutashvili
A New Approach to Analysing Fuzzy Data and Decision-Making Regarding the Possibility of 299
Earthquake Occurrence
Jina Gachechiladze, Tamaz Gachechiladze, Irina Khutsishvili
Investigation of the Heat Index in Georgia based on the Most Typical Fuzzy Expected Values 302
Marina Fandoeva, Tatiana Kiseliova, Anna Sikharulidze
The Combined Decision Making Method based on the Statistical and Fuzzy Analysis 309
Irina Khutsishvili
A Fuzzy Identification Problem for the Stationary Discrete Extremal Fuzzy Dynamic System 323
Gia Sirbiladze
Data Unification Algorithm for Representing Incomplete and Indefinite Information in the 348
Medical Expert System
Manana Khachidze, Maia Mikeladze
Problem of Determining Optimal Lot Size with Respect to the Production, Storage, and Quality 353
Criteria
Shorena Okujava, Tamar Asatiani, Ia Giashvili, Manana Magradze
A Recursion Forms and their Verification by using the Inductive Methods 357
N. Archvadze, M. Pkhovelishvili, L. Shetsiruli, M. Nizharadze
The Dominance Concept of Dempster-Shafer (D-S) Belief Structure in the Modeling Decisions 362
Zakaria Karsaulidze, Gia Sirbiladze, Bezhan Ghvaberidze
Fixing Variable Values in the Capacitated Facility Location Problem through Pairing and 375
Surrogate Constraint Analysis
Maria A. Osorio, Abraham Sanchez
The Two-Stage Model for Order Selection and Operation Sequencing 401
Anton Lominadze, Nodar Lominadze
Analog Filter Group Delay Optimization. Efficiency Evaluation of Optimization Methods 421
Karel Zaplatilek
Plenary Lecture 1
Abstract: The Evolutionary Computation covers sets of algorithms inspired by the theory of the species evolution.
The interconnection between the Artificial Intelligence techniques such as Genetic Algorithms and Artificial Neural
Networks for the creation of evolutionary systems is applied in the resolution of various types of problems in particular
in the optimization processes. The objective of this union is to combine symbolic and conexionist systems in order to
optimize the description of the universe of discourse represented by logic functions. We focus on the study of the
combination of Artificial Neural Networks with Genetic Algorithms to investigate the descriptive power of the Extended
Logic Programming in order to qualify the description of the universe of discourse or theories expressed by logic
mathematical functions. With this theories and methodologies the goal is to get the best value of the quantification
theory which corresponds to the best logic mathematical functions and try to optimize it.
Brief Biography of the Speaker: Jorge Ribeiro was born in 1975, in Braga, Portugal and is Assistant Professor of
the School of Technology and Management Polytechnic of the Institute of Viana do Castelo - Portugal. Is teaching in
the Artificial Intelligence, Systems Integration and Enterprise Information Systems field. Is member of the Artificial
Intelligence Group of the Informatics Department of the University of Minho - Portugal. He received MSc degree in
computer science (2002) from the University of Minho-Portugal. His PhD dissertation (at the Department of Electronic
and Computation of the University of Santiago de Compostela, Spain.) addresses the optimization of logic
mathematical functions applying evolutionary systems in order to maximize the quality-of-information. He has been
an author and co-author of some papers in the field of Data Mining, Software Engineering, Knowledge
Representation and Evolutionary Systems.
Plenary Lecture 2
Visualizing Programs
Abstract: During their live cycle, programs have a tendency to grow and, by the way, to increase their complexity,
thus becoming harder and harder to read, to understand and to maintain. But understanding programs is necessary
for their continuous evolution, their maintenance and debugging.
Various tools, such as metrics, abstract interpretation, model checking, code coverage, program slicing, data-flow
analysis, pointer analysis, call graphs, data-flow graphs and dependency graphs (to name but a few), have been
developed to help programmers in their understanding of the structure and functioning of programs. In this lecture we
will present various ways of displaying graphs and focus on the impact of visualization of such static and dynamic
data dependencies in the context of program maintenance. We will examine several prototypes we have developed in
our laboratory, detailing especially the different visualizations we propose, which are aimed to minimize the
conceptual overload in order to allow users to deal with hard to understand or buggy programs.
Brief Biography of the Speaker: Harald Wertz was born in 1947 in Friedrichshafen, Germany, and is living in
France since 1971. He is Professor of Computer Science at the Universite Paris 8, France, since 1978, and he is a
founding member of the Franco-Georgian Institute. He loves to teach introductory programming classes, artificial
intelligence and software engineering. His research focuses mainly on the intersection between artificial intelligence
and software engineering: finding computational methods helping to understand, debug and meaningfully represent
programs. His doctoral thesis, at the University of Paris 6, described a Lisp-system able to automatically debug
programs written by novice programmers. His ‘these d’etat es science’, at the Universite de Vincennes, described the
structure, implementation and use of an integrated and incremental programming environment, which included (then)
novel features such as automatic documentation, automatic construction of outstanding task lists, on the fly
correction, reverse execution and executable program annotation. He is the author of three books, several book
chapters and some seventy papers.
Plenary Lecture 3
Abstract: The measurement information systems make use of different converters whereby the link between the
signal converters and the system bus impose structural and algorithmic methods for increasing accuracy.
The simulation methods are applied for measurement of powers and energies in electrical networks using instrument
transformers. The transformers allow the measurement of the effective value but, in certain cases, one can note
errors in the measurement of the electric power and energy.
In most practical cases a current transformer consists of a traditional instrument with magnetic core. Accuracy
specifications of these devices are generally assured under sinusoidal conditions. When distorted waveforms are
dealt with, CTs can introduce large uncertainties, thus leading to gross measurement errors.
A variety of techniques is reported for improving the performance of instrument transformers. In many cases they
refer to hardware circuits for electronically compensated devices. More recently, several different software
compensation techniques have been introduced for limiting the errors. Following a similar approach, digital
compensation methods have been proposed to significantly reduced hysteresis and eddy currents effects.
In the present paper, we present different models of current transformers and computer methods for the simulation of
characteristics and evaluation of measurement errors, with the goal of increasing the measurement accuracy.
Brief Biography of the Speaker: Costin Cepisca was born in Bucharest, Romania, on May 21, 1949. He received
the degree in electrical engineering (1972) and the Ph.D. (Dr.ing.) from Bucharest Polytechnic Institute in 1983.
He is currently Professor of measurement systems at the POLITEHNICA University of Bucharest, Head of Research
Centre for Metrology and Measurement Systems and former Vice-Dean of Faculty of Electrical Engineering.
His present research interest includes the sensor interface systems, analogue circuit design and signal processing,
measurement theory and power quality. He has published more 600 technical papers and 60 books and has been
involved in numerous government and industrial projects in area of measurements and instrumentation.
Plenary Lecture 4
Abstract: This paper relates the theory of optimal control to the Finsler geometry and to the geometric dynamics.
Section 1 shows how we build a Finsler structure starting from an optimal control problem. Section 2 studies a
dynamic system evolving over time and controlled through a Finsler control. In order to choose properly the controls,
we use a payoff functional and the Pontryaguin Maximum Principle. Section 3 transforms a controlled evolution
system in a Finsler geometric dynamics. Section 4 introduces the idea of the controlled Finsler gradient flows in
optimization problems. Section 5 studies the Finsler optimal control attached to Rosenbrock function. Section 6
formulates a Finsler affine regulator problem. Section 7 describes a Finsler optimal growth model.
Brief Biography of the Speaker: Important Career Positions: Dean, Director, Chair, Full Professor 1990-, University
Politehnica of Bucharest, Department of Mathematics.
Number of PhD Students: 25 in due time and 13 Doctors in Mathematics.
Membership of Associations: AMS, 1987; Tensor Society, 1985; Balkan Society of Geometers, President, 1994;
Publications: over 40 books; 200 papers; 200 communications.
Honors: D. Hurmuzescu Prize, Romanian Academy, 1985; Award MEI, 1988; Correspondent Member, Academia
Peloritana, Messina, 1997; Titular Member, Academy of Romanian Scientists, 2007; Honorary Member, World
Scientific and Engineering Academy and Society, 2008-;
Organizer: The International Conference of Differential Geometry and Dynamical Systems, University Politehnica of
Bucharest, October 5-7, 2007; 7th WSEAS International Conference on Systems Theory and Scientific Computation
(ISTASC'07), Vouliagmeni Beach, Athens, Greece, August 24-26 (2007); European Computing Conference,
Vouliagmeni Beach, Athens, Greece, September 24-26, 2007; 12th WSEAS International Conference on Applied
Mathematics, Cairo, Egypt, Dec. 29-31, 2007; 7th WSEAS International Conference on Circuits, Systems,
Electronics, Control and Signal Processing, Cairo, Egypt, Dec. 29-31, 2007; Chair-Committee: American Conference
on Applied Mathematics (Math'08) and Management, Marketing and Finances (MMF'08), Cambridge, Massachusetts,
USA, March 24-26, 2008; International Program Committee: The Applied Computing Conference (ACC-08), Istanbul,
Turkey, May 27-30, 2008.
Fields of Interest: Differential Geometry, Optimizations on Riemannian Manifolds, Magnetic Dynamical Systems,
Geometric Dynamics.
Plenary Lecture 5
Abstract: Data flow reduction in digital instrumentation usually comes with the benefit of statistical processing for
noise cancellation and preservation of signals correlation parameters among an overall compression of information.
Over-sampling techniques offer the unique characteristic of noise shaping in analog to digital converters providing
additional accuracy for measurements with the cost of local signal processing for decimation of data.
We present several techniques of data aggregation regarding sigma-delta modulators used in energy counters and
power quality analyzers performing unconventional statistical data processing and data aggregation bought in time
and frequency domains. Simulations for code and time interpolation data reduction have been performed for primary
data flow of the measurement chain and for data collection aggregation for compact power quality parameters
extraction.
Brief Biography of the Speaker: Sorin Dan Grigorescu was born in Bucharest, Romania on June 8, 1958. He
received the degree in electronics and telecommunications (1984) and the Ph.D. (dr. eng.) from the Bucharest
Polytechnic Institute in 1996.
He performs research and teaching as Professor of virtual instrumentations and distributed measurement systems at
the POLITEHNICA University of Bucharest. His research field includes signal processing, monitoring complex
systems, power quality and transducers.
He has published more than 200 scientific papers and 20 books, being the head of research team for several
governmental and industrial projects in the fields of instrumentation, power quality and integrated control of the drilling
rigs.
Plenary Lecture 6
Abstract: Regardless of their increasing number and diversity, the capacities of Neural Network (NN) models still
remain far behind the ones biological systems can exhibit when faced to changing environments or other complex
processes. As an attempt to better understand why, we propose to investigate the variations of a few NN algorithms
in the theoretical framework of Darwinian evolution in order to favor the emergence of more global models through
gradual adaptive developmental steps.
First, we explore the possibility to achieve a more general conception of learning and training methodology, detached
from specialized NN models. Then, with the ultimate goal to bypass human bias constraints in data acquisition, we
apply NNs to the automatic categorization of natural language data without prior knowledge.
To this end, we argue that enhanced plasticity and reorganization capabilities are necessary for NNs in order to be
able to detect and structurally integrate variations in the data space. A first step is to model and simulate the dynamic
character of the “biological” learning structures and processes as well as their evolution over time. We propose a dual
architecture, where two –possibly identical– NNs collaborate, one learning to control the efficiency of the other. This
way, a reflexive loop of self-supervision can be achieved where one NN learns to tune the configuration parameters
(wiring, growth, learning rate, etc.) of the other through automated trial and error sessions. A further step is the use of
data driven programming combined with error measures in the self-supervision loop to create a self-observing
retroactive loop in order to analytically develop an active, event guided, learning. The previously mentioned dual
architecture can then be used to learn to extract and apply characteristic learning features of other NN models. NNs
would thereby, in response to environmental changes, put into practice their acquired adaptive developmental
capabilities to generate the appropriate variations, both at the architectural and procedural level.
We will also distinguish different scales of structural variations in the natural – and mostly biological – world in order
to illustrate emergent steps in the evolution of developmental strategies, similar to those nature has selected.
Brief Biography of the Speaker: Jean-Jacques Mariage was born in Saisseval, France, in 1953. He is Associate
Professor of Computer Science at the University of Paris 8 since 2001, where he joined the Artificial Intelligence
Laboratory. He teaches at the Franco Georgian Institute since 2006. His teaching activity involves Artificial
Intelligence, programming languages and computer network engineering. His current research addresses the
integration of a self-observing retroactive loop in unsupervised Neural Network (NN) models, applied to the automatic
categorization of natural language data without previous knowledge. To this end, his interest focuses on the modeling
and simulation of the developmental dynamics of adaptive encoding structures as found in biological systems. His
areas of interest include learning, memory, evolution, NN algorithms, genetic algorithms, evolutionary programming,
artificial life as well as the development, replication and adaptation of biological encoding structures.
Plenary Lecture 7
Abstract: This work deals with the problem of the Stationary Discrete Extremal Fuzzy Dynamic System (SDEFDS)
identification and briefly discusses the results developed by G. Sirbiladze. The fuzzy process with possibilistic
uncertainty, the source of which is expert knowledge reflections, is constructed. The dynamics of SDEFDS is
described. Based on the fuzzy-integral model, methods and algorithms are developed for identifying the transition
operator of the stationary discrete extremal fuzzy dynamic system. The SDEFDS transition operator is restored by
means of expert data with possibilistic uncertainty, the source of which is expert knowledge reflections on the states
of SDEFDS in the extremal fuzzy time intervals.
The regularization condition for obtaining a quasi-optimal estimator of the transition operator is represented by the
theorems. The corresponding calculating algorithm is provided. The results obtained are illustrated by the example in
the case of a finite set of SDEFDS states.
Brief Biography of the Speaker: Dr. Gia Sirbiladze is a full professor at the Department of Computer Science of
Faculty of Exact & Natural Sciences of Iv. Javakhishvili Tbilisi State University, Georgia. He received his Ph.D.
degree in 1991 from the Computational Mathematics Institute of the Georgian Academy of Science. He received his
D. Sci. degree from the same institute in 2005. His scientific interests include areas such as Intelligent Fuzzy
Technologies and General Systems, Fuzzy Technologies in Decision-making Support Systems, Fuzzy Extremal
Dynamic Systems - Control, Filtration and Identification, Fuzzy Discrete Optimization Problems and Modeling
Decisions. Dr. Gia Sirbiladze has published 54 scientific papers on the above-listed topics. He is an author of one
monograph on Decision Making Problems in General Environment. Dr.Gia Sirbiladze has participated in many
scientific conferences, including plenary speeches on WSEAS conferences. Dr.Gia Sirbiladze is a member of the
National Union of Mathematicians in Georgia. He serves as a reviewer for Mathematical Reviews. He has reviewed
papers for more then 15 international and local journals and conferences. He serves as Information Technology
expert for Georgian National Scientific Fund. Dr.Gia Sirbiladze has participated in several national and international
research projects.
Plenary Lecture 8
On the Optimality of a General Production Lot Size Inventory Models with Variable
Parameters
Abstract: A general production lot size dynamic inventory model with deteriorating items for which the rates of
demand, production, deterioration as well as the cost parameters are arbitrary and known functions of time is
considered in this paper. Shortages are allowed but are partially backordered. Both inflation and time value of money
are taken into account. The objective is to minimize the total net inventory cost. The relevant model is built, solved
and some main results about the uniqueness and the global optimality of this solution, with the use of rigorous
mathematical methods, are obtained. An illustrative example is provided.
Brief Biography of the Speaker: Degrees: B.Sc. in mathematics (Probability and Mathematical Statistics Section)
Damascus University 1971-High Diploma in pure mathematics, Damascus University 1979 - Studying several post
graduate courses in Optimization, Probability and Statistics Brussels University 1980 - Dr. of Science in applied
mathematics (Belgian Ph.D in applied mathematics- OR oriented) with honors, University of Brussels 1983.
Positions: Professor in King Saud University, College of Science, Department of Statistics and Operations Research
since May 2005.,Associate Professor King Suad University, College of Science, Department of Statistics and
Operations Research, 1998 -Assistant Professor, King Saud University, College of Science, Department of Statistics
and Operations Research, From 1983 to 1998- Demonstrator in Mathematic Department in Damascus University
1979.
Professional Experience: He has about 27 years academic teaching experience in Operations Research, Statistics,
and Mathematics, supervising several M.Sc. and PhD thesis., Main contributor in the design and development of the
B.Sc. and M.Sc. and PhD Programs in Operations Research And Statistics in the Department of Statistics and
Operations Research in King Saud University (Riyadh - Saudi Arabia), Member in the editorial board of “Journal of
Scientific Inquiry” , Acting as a referee for more than 15 specialized and leading international journals (more than 20
papers per year) . He contributed in many Local and International Scientific Conferences and Symposiums. He
Contributed in giving consultations in solving local real problems in Saudi Arabia using the OR techniques.
His Research Interests: are in Applied Mathematic (Operations Research Oriented). In particular, Optimal Search
problems where he has more than 10 research papers. Recently he research turns to Inventory Control problems
with more than 35 publications in a scientific and leading journals.. He also authored 4 books in Fundamentals of
Operations Research, Inventory Control, Game Theory, and Integer Programming.
Authors Index