Abstract: One of the basic engineering problems of the automatic control systems synthesis for steam turbines is systems quality indices optimization task. The features of such task are defined by plenty of control systems structural parameters, complication of quality indices formalization and calculation, the systems models high order. The greatest difficulty of the control system synthesis is optimization models and methods design. The paper purpose is to develop models and methods for optimum design of the intelligent automatic control systems of atomic power station steam turbine using genetic algorithms. The steam turbine automatic control system is applied to stabilize turbine rotor frequency with high precision. The intelligent steam turbine control system includes a steam turbine, frequency sensor, intelligent frequency regulator, electro-hydraulic automatic drive and turbine adjusting valve. Input signals on an automatic drive can be given by the steam pressure regulator through the turbine management mechanism or by the electric power regulator. Assumptions substantiated to model the automatic control system and automatic drive as executive link of control system. The automatic drive diagram of principle includes electro-hydraulic transformer, sleeve valve, servomotor, position sensors and electronic part. In the paper were built the mathematical models of automatic drive and steam turbine, the models permanent parameters values were calculated, the mathematical models of the automatic control system were developed in state space with the intelligent frequency regulator, the regulator parameters were optimized with system quality indices using genetic algorithms.

Keywords: automatic control system, automatic drive, steam turbine, intelligent regulator, genetic algorithms.

ACM Classification Keywords: G.1.6 Optimization - Nonlinear programming

Link:

OPTIMAL DESIGN OF INTELLIGENT CONTROL SYSTEMS OF STEAM TURBINE USING GENETIC ALGORITHMS

Khrystyna Fedyanyna, Ievgeniia Kucher, Valery Severin

http://foibg.com/ibs_isc/ibs-13/ibs-13-p15.pdf

Abstract: The paper is devoted to the development of a perspective concept of atomic station power block intelligent automatic control systems synthesis on the basis of mathematical models and numeric methods of vector optimization of systems quality indexes using genetic algorithms. The methods for calculation of direct quality indexes and improved integral quadratic estimates have been created. The step-by-step principle of transition to the domain of system stability has been based. There have also been suggested vector objective functions including stability conditions and taking into consideration quality indexes priorities. The reliable genetic algorithms for vector objective functions optimization have been suggested. Mathematical models in the state space for intelligent automatic control systems of nuclear reactor and steam generator have been worked out. The quality indexes optimization of power block intelligent control systems has been carried out, which allowed to estimate various controller types efficiency.

Keywords: automatic systems, intelligent control, optimal synthesis, nuclear reactor, steam generator, genetic algorithms.

ACM Classification Keywords: G.1.6 Optimization - Nonlinear programming

Link:

OPTIMAL SYNTHESIS OF INTELLIGENT CONTROL SYSTEMS OF ATOMIC POWER STATION USING GENETIC ALGORITHMS

Seyed Mojtaba Jafari Henjani, Valery Severin

http://foibg.com/ibs_isc/ibs-13/ibs-13-p14.pdf

Abstract: Methods for calculation of quality indexes for automatic control systems are presented. For the optimization of quality indexes defined only in a stability domain a vector objective function of varied parameters of the system is proposed. The stepwise principle of successive satisfaction of constraints for the passage into the definition domain of quality indexes is considered, as well as a rational mechanism of its realization in the form of the priority optimization of the vector objective function. For the optimization of the vector objective functions genetic algorithms as vector optimization methods are presented. Their application allows one to steer the optimization process from any initial point of the space of varied parameters into the stability domain of the system and to find the optimum of the quality indexes in this domain. The efficiency of the proposed application of vector genetic algorithms for the quality indexes optimization is confirmed by computational experiments.

Keywords: genetic algorithms, vector optimization methods, automatic control systems, quality indexes.

ACM Classification Keywords: G.1.6 Optimization - Nonlinear programming

Link:

APPLICATION OF GENETIC ALGORITHMS TO VECTOR OPTIMIZATION OF THE AUTOMATIC CONTROL SYSTEMS

Valery Severin

http://foibg.com/ibs_isc/ibs-13/ibs-13-p13.pdf

Abstract: The article presents proposition of the assist pilot system. The system helps pilot with correct decision of cross country flights. Presentation of specially calculated information stored by the system delivered to pilot could help him with his decision about correct directions of crossing.

Keywords: Knowlage base, Data mining

ACM Classification Keywords: E. Data, E.5 Files, E.5.6. Sorting/searching

Link:

THERMALS MAP – ASSIST FLIGHT SYSTEM

Arkadiusz Rzucidło

http://foibg.com/ibs_isc/ibs-13/ibs-13-p12.pdf

Abstract: The article is described architecture and functionality of development support system "OntoIntegrator” for NLP.

Keywords: Natural language processing, ontological models, program tools for NLP

ACM Classification Keywords: H.3.1.Information storage and retrieval: linguistic processing

Link:

THE DEVELOPMENT SUPPORT SYSTEM "ONTOINTEGRATOR" FOR LINGUISTIC APPLICATIONS

Olga Nevzorova, Vladimir Nevzorov

http://foibg.com/ibs_isc/ibs-13/ibs-13-p11.pdf

Abstract: It is known, that person’s thinking is inaccessible to studying by direct physical and psychological methods. In this case it is necessary to have indirect ones. Computers do not understand the psychological description and formalization of thinking mechanisms. Algebra logic analysis of natural language and person’s thinking plays an important role for development of logic mathematics and its applications in artificial intelligence. Only axiomatic method works in this situation. On the basis of axioms’ system we can propose an approach that helps to investigate the structure and properties of objects. The main problem of formal studying of a natural language is shortage of the mathematical apparatus. The axiomatic description of logic mathematics' objects requires preliminary realization of constructive logical tools, which subsequently become a subject of the axiomatic analysis. The paper is devoted the algebra of ideas to axiomatic construction. The carrier of this algebra is naturally interpreted as the set of intelligence ideas (thoughts, concepts and, in general, any subjective conditions of the person). There are devised some methods for application of proposed formal apparatus. Simultaneously with algebra of ideas formal introducing there is considered its intentional interpretation.

Keywords: a predicate, algebra of ideas, artificial intelligence, Cartesian set, algebra of predicates, algebra of single k-dimentional first order predicates.

ACM Classification Keywords: I.2 Artificial intelligence - Natural Language Processing

ALGEBRA LOGIC APPROACH TO PERSON’S THINKING MECHANISMS FORMALIZATION

Olga Kalinichenko

http://foibg.com/ibs_isc/ibs-13/ibs-13-p10.pdf

Abstract: Challenges of knowledge society, knowledge economics change models of management. Work trends, organisational structures are changing. Most of knowledge management processes one way or another are connected with main elements of information management: information technologies and information systems. Knowledge management is new branch of management, the main purpose of that is bigger effectiveness of business applying synergy of humans, processes and technologies. Knowledge management helps for employees single-mindedly to create, share, collect, keep and practice the knowledge. Fundamentals of knowledge management could be used for organisation of telework of disabled persons, solving economical problems of state. Knowledge management could help to increase productiveness of employees, expanding sources of reachable for them knowledge. Improving shapes of disabled persons’ education the very important is to pay attention at how educational knowledge is implemented practically. Knowledge management should encourage employees to integrate such educational processes together with their work methods and to apply results of educational processes in their daily activities. For purposeful and beneficial education there should be forecasted what knowledge is the most important. Authors of article analyses the application of knowledge management methods in structure of professional rehabilitation of disabled persons, organising telework of disabled persons. During such process new knowledge should be used not only by disabled persons, but both represents of governmental institutions and consultants.

Keywords: knowledge society, disabled persons, telework.

ACM Classification Keywords: K.4.3. Computers and society - Computer-supported collaborative work

Link:

ORGANISATION OF DISABLED PERSONS’ TELEWORK AS PART OF PROCESS OF KNOWLEDGE MANAGEMENT

Tatjana Bilevičienė, Eglė Bilevičiūtė

http://foibg.com/ibs_isc/ibs-13/ibs-13-p09.pdf

Abstract: This work presents the using of the chosen neural networks for the data classification. The Statistic Neural Networks has been used for the rating. The neural networks have been successfully applied for building one decision support systems for solving managerial problem.

Keywords: artificial intelligence, neural networks, classification, decision support

ACM Classification Keywords: I. Computing Methodologies, I.2 Artificial Intelligence

Link:

A NEURAL NETWORKS APPLICATION IN DECISION SUPPORT SYSTEM

Galina Setlak, Wioletta Szajnar, Leszek Kobyliński

http://foibg.com/ibs_isc/ibs-13/ibs-13-p08.pdf

Abstract: In this paper the review of applications of data mining is presented. The algorithm of the induction across decision trees that was proposed for classifying large set of data is described in the part 2 of the paper. A short characteristic of the proposed algorithm is included in the part 3. In the part 4 we present characteristics of R-environment and the implementation of the algorithm. The method of conducting the computational experiments and the obtained results are presented in the part 5. The last part of the paper contains conclusions and further perspectives.

Keywords: data mining, classification, critical defects, induction of decision trees formatting rules

ACM Classification Keywords: I. Computing Methodologies, I.2 Artificial Intelligence, J. Computer Applications J.6 Computer-aided engineering,

Link:

THE APPLICATION OF DATA MINING METHODS TO THE CLASSIFICATION OF FINISHED PRODUCTS

Monika Piróg-Mazur?, Galina Setlak, Wioletta Szajnar, Tomasz Kożak

http://foibg.com/ibs_isc/ibs-13/ibs-13-p07.pdf

Abstract: ARES System is an application dedicated to data analysis, data exploration and knowledge discovery. This versatile tool offers rough set based methodology as well as emerging patterns and Support Vector Machine algorithms to be applied to input data presented as an information system. Many analytical capabilities provided with an user-friendly, intuitive graphical interface and XML based support for input/output operations make ARES System a challenging and promising software for application in scientific research and didactic work. An unique feature of ARES System is its ability of identifying improper objects within information system by employing credibility coefficients. The credibility coefficients attempt to assess a degree of typicality of each object in respect to the whole information system.

Keywords: information system, classification, rough sets, emerging patterns, SVM, credibility coefficients.

ACM Classification Keywords: H.1.1 Systems and Information Theory

Link:

ARES SYSTEM - INTEGRATION OF ANALYSIS METHODOLOGIES

Roman Podraza, Mariusz Kalinowski

http://foibg.com/ibs_isc/ibs-13/ibs-13-p06.pdf

Abstract: Identification and prediction problem of nonlinear time-series generated by discrete dynamic system is considered via Kernel Method approach. A unified approach to recurrent kernel identification algorithms design is proposed. In such a way a recurrent modification of initial Kernel Method with growing windows is considered. In order to prevent the model complexity increasing under on-line identification, the reduced order model kernel method is proposed and proper recurrent identification algorithms are designed along with conventional regularization technique. Such an approach leads to a new type of Recursive Least-Square? Kernel Method identification algorithms. Finally, the recurrent version of Sliding Window Kernel Method is also developed along with suitable identification algorithms. The proposed algorithm has tracking properties and may be successfully used for on-line identification of nonlinear non-stationary time-series.

Keywords: identification, kernel methods, machine learning, nonlinear model, prediction, recurrent least-squares, support vector machine, time-series

ACM Classification Keywords: G. Mathematics of Computing: G.1 Numerical Analysis: Least squares approximation, nonlinear approximation, G.3 Probability and Statistics: Time series analysis

Link:

KERNEL-BASED METHODS FOR NON-STATIONARY TIME-SERIES IDENTIFICATION AND PREDICTION

Leonid Lyubchyk, Vladyslav Kolbasin

http://foibg.com/ibs_isc/ibs-13/ibs-13-p05.pdf

Abstract: The authors have undertaken a research task with a view to apply of evolutionary algorithms and artificial neural network to design of the course of a gas nitriding process, which guarantees to obtain the expected hardness profile in the nitrided layer. The gas nitriding process is widely used in industry in order to improve the functional properties of machine and tool parts. First of all, the artificial neural network was trained dependences between physical properties of steel and process parameters in relation to hardness profile, formed in gas nitriding process. Those data was collected experimentally. Such trained neural network was used as a mathematical model for the design approach of gas nitriding process. The aim of such designing was to predict the parameters of nitrinding industrial process, in which the required hardness profile of nitriding layer will be obtained. Prediction of manufactory conditions was realized with the help of evolutionary algorithms. In the approach developed, each chromosome includes encoded parameters of a single steel nitriding process. For each chromosome, a steel hardness profile related to it is determined by the model which is represented by neural network. On the further step, the chromosomes undergo a selection, a modification with the aid of crossing mutation operations and promotion to the next population. In this manner, by approach of a directed evolution of steel nitriding parameters, one is selected for which the hardness profile matches the best to profile sought.

Keywords: evolutionary algorithm, neural network, designing of nitriding process, nitriding layer, hardness profile, model, manufactory conditions prediction.

ACM Classification Keywords: D.2.2 Design Tools and Techniques - Evolutionary prototyping

Link:

THE APPLICATION OF ARTIFICIAL NEURAL NETWORKS AND EVOLUTIONARY ALGORITHM FOR THE DESIGNING OF GAS NITRIDING PROCESS

Mateusz Kosikowski, Zbigniew Suszyński, Roman Olik, Jerzy Ratajski, Tomasz Suszko

http://foibg.com/ibs_isc/ibs-13/ibs-13-p04.pdf

Abstract: New non-conventional system of the computational intelligence is proposed. It has growing structure similar to the Cascade-Correlation? Learning Architecture designed by S. E. Fahlman and C. Lebiere but differs from it in type of artificial neurons. Quadratic neurons are used as nodes in introduced architecture. These simple elements can be quickly adjusted using high-speed learning procedures. Proposed approach allows to reduce time required for weight coefficients adjustment and to decrease training dataset size in comparison with conventional neural networks. Also on-board realization of quadratic neuron is quite simple and therefore implementation of entire cascade architecture in hardware is very easy.

Keywords: artificial neural networks, constructive approach, quadratic neuron, real-time processing, online learning.

ACM Classification Keywords: I.2.6 Learning – Connectionism and neural nets.

Link:

THE CASCADE GROWING NEURAL NETWORK USING QUADRATIC NEURONS AND ITS LEARNING ALGORITHMS FOR ON-LINE INFORMATION PROCESSING

Yevgeniy Bodyanskiy, Yevgen Viktorov, Iryna Pliss

http://foibg.com/ibs_isc/ibs-13/ibs-13-p03.pdf

Abstract: Theory of Infos is a part of the General Information Theory. Main features of the Theory of Infos are outlined in the paper. Further investigation is discussed.

Keywords: General Information Theory, Theory of Infos.

ACM Classification Keywords: A.1 Introductory and survey

Link:

THEORY OF INFOS

Krassimir Markov, Krassimira Ivanova, Ilia Mitov

http://foibg.com/ibs_isc/ibs-13/ibs-13-p01.pdf

Abstract: A Bulgarian research project funded by the Bulgarian National Science Fund under the thematic priority Information and Communication Technologies, contract D002-308/19.12.2008 is presented in this article. The main goal and tasks are outlined. Some already achieved results are pointed.

Link:

AUTOMATIC METADATA GENERATION FOR SPECIFICATION OF E-DOCUMENTS – THE METASPEED PROJECT

Juliana Peneva, George Totkov, Peter Stanchev, Elena Shoikova

http://foibg.com/ibs_isc/ibs-14/ibs-14-p14.pdf

Abstract: EulerPathSolver? is a new application, that meets eMathTeacher specifications and simulates Fleury’s algorithm execution. The application runs in a Java Web Start Window and features an animation of the algorithm code, a framework working panel showing the algorithm structures and allowing their manipulation, Pop-up questions, language selector and, save/load options together with the interactive simulation within automatic correction of the user’s inputs. It has been designed with the main purpose of supporting active learning as well as being a good aid for teachers when explaining the algorithm process. EulerPathSolver? enhances dramatically the old Fleury’s Algorithm tutorial designed by the authors and will be a great partner when learning Fleury’s Algorithm.

Keywords: eMathTeacher, eLearning, active learning, interactive Java applications, discrete mathematics learning, algorithm visualization.

ACM Classification Keywords: K.3.1 Computers and Education: Computer Uses in Education – computerassisted instruction (CAI), distance learning. K.3.2 Computers and Education: Computer and Information Science Education – computer science education, self-assessment. G.2.2 Discrete Mathematics: Graph Theory – graph algorithms, path and circuit problems.

Link:

EULERPATHSOLVER: A NEW APPLICATION FOR FLEURY’S ALGORITHM SIMULATION

Gloria Sánchez–Torrubia?, Carmen Torres–Blanc?, Leila Navascués-Galante?

http://foibg.com/ibs_isc/ibs-14/ibs-14-p13.pdf

Abstract: This paper presents an adaptable InfoStation-based? multi-agent system facilitating the mobile eLearning (mLearning) services provision within a University Campus. The network architecture is presented, and the interactions between the various components within the architecture during mLearning service provision (mLecture, mTest) are presented. System implementation approaches are also considered, with particular attention paid to the creation of user profiles and service profiles for the personalization and contextualization of the presented services.

Keywords: InfoStations?, Intelligent Agents, Multi-agent System, mLearning, CC/PP-UAProf, OWL-S.

Link:

ADAPTATION FOR ASSIMILATION: THE ROLE OF ADAPTABLE M-LEARNING SERVICES IN THE MODERN EDUCATIONAL PARADIGM

Damien Meere, Ivan Ganchev, Stanimir Stojanov, Máirtín O’Dróma?

http://foibg.com/ibs_isc/ibs-14/ibs-14-p12.pdf

Abstract: An approach for indirect spatial data extraction by learning restricted finite state automata from web documents created using Bulgarian language are outlined in the paper. It uses heuristics to generalize initial finite-state automata that recognizes only the positive examples and nothing else into automata that recognizes as larger language as possible without extracting any non-positive examples from the training data set. The learning method, program realization and experiments are presented. The investigation is carried out in accordance and following the rules of EU INSPIRE Network.

Keywords: Automatic Data Extraction, Restricted Finite State Automata, Web Documents, Indirect Spatial Data, INSPIRE network.

ACM Classification Keywords: H.2.8 Database Applications - Data mining; F.1.1 Models of Computation – Finite State Automata

Link:

INDIRECT SPATIAL DATA EXTRACTION FROM WEB DOCUMENTS

Dimitar Blagoev, George Totkov, Milena Staneva, Krassimira Ivanova, Krassimir Markov, Peter Stanchev

Abstract: This article is devoted to the questions of prognostication of efficiency medical and prophylactic measures at renewal of the broken equilibrium of human organism by vegetable medications. The methods of modeling of different processes are considered in the article from position of systems analysis. There was chosen Markov processes theory for description of renewal of the broken homoeostasis process, which rotined high correlation between clinical supervisions and forecast process.

Keywords: casual process, Markov chains, medical and prophylactic measures, MAIS, MFPS.

ACM Classification Keywords: I.6 Simulation and modeling, I.6.3 Applications; J.3 Life and medical sciences – Medical information systems

PROGNOSTICATION OF EFFICIENCY OF MEDICAL AND PROPHYLACTIC MEASURES AT DIFFERENT HOMOEOSTASIS VIOLATION OF HUMAN ORGANISM BY MARKOV PROCESSES THEORY

Boris Samura, Anatoly Povoroznuk, Olga Kozina, Elena Visotskaja, Elena Chernykh, Nikolay Shukin, Andrei Porvan

Abstract: Based on several results related to studies on emotions, we suggest that the process of emotionrecognition is assisted by some internal structure of the cognitive images of emotions, which are at different levels of knowledge representation. We concede that the main proposed in psychology models are in correspondence with these levels and in this sense - complementary. We summarize the state-of-the-art of machine emotion recognition with regards of the used psychological models of emotions. In order to discover amelioration sources of multimodal machine emotion recognition, we propose a scheme of the cognitive process, based on gradual levels of representation. The proposed scheme shows several "strategic" differences with the architectures used in machine emotion recognition. We discuss the questions related to recognition, assisted by two levels of representation that we called "perceptual" and "conceptual".

ACM Classification Keywords: I.2 Artificial Intelligence, 1.2.0.Cognitive simulation

Link:

MULTI-MODAL EMOTION RECOGNITION – MORE "COGNITIVE" MACHINES

Velina Slavova, Hichem Sahli, Werner Verhelst

http://foibg.com/ibs_isc/ibs-14/ibs-14-p09.pdf

Abstract: The development of a methodology and tools for an automatic extraction of metadata for digital objects deployed in various subject repositories is a potential research issue. This paper presents a general overview of topics concerning the building of repositories and the applied metadata schemas with respect to the objectives of METASPEED project. The main goal of this project is to investigate and create technologies, methods and tools for automatic generation of metadata thus facilitating the proper specification of documents with different eformat, content and location.

Key words: Digital object repository, metadata schema

Link:

DIGITAL OBJECTS – STORAGE, DELIVERY AND REUSE

Juliana Peneva, Stanislav Ivanov, Filip Andonov, Nikolay Dokev

http://foibg.com/ibs_isc/ibs-14/ibs-14-p08.pdf

Abstract: In this paper an analysis of possibilities offered by virtual environments for accomplishing attacks to and within it, is made. Main techniques for accomplishing an attack to virtual environment are pointed and real virtual attacks and successful virtual attacks are examined. An analysis of accomplished attacks is made and percent distribution of real virtual attacks and successful virtual attack is graphically illustrated. Respective assessments and recommendations for future investigation are made.

Keywords: Virtual Machine Environment, Virtual Environment, Attack/Attack tools, Defense/Defense tools, Malicious Code

ACM Classification Keywords: D.4.6 Security and Protection: Invasive software (e.g., viruses, worms, Trojan horses)

Link:

ANALYSIS OF MALICIOUS ATTACKS ACCOMPLISHED IN REAL AND VIRTUAL ENVIRONMENT

Dimitrina Polimirova, Eugene Nickolov

http://foibg.com/ibs_isc/ibs-14/ibs-14-p07.pdf

Abstract: Admission control is a key issue for quality of service (QoS) provisioning in both wired and wireless communication networks. The call admission control (CAC) algorithm needs to know the source traffic characteristics and the required performance in order to determine whether the connection can be accepted or not and , if accepted, the amount of network resources to allocate. In this paper, we determine the CAC threshold value in case streaming homogeneous ON-OFF traffic flow is considered. An analytical method for packet loss probability evaluation is proposed and numerical examples are presented.

Keywords: streaming traffic, ON-OFF traffic model, call admission control, packet loss probability

ACM Classification Keywords: C.2.1 Network architecture and Design – Wireless communication, C.2.5 Local and Wide-Area? Networks – Access schemes

Link:

PERFORMANCE ANALYSIS OF CALL ADMISSION CONTROL FOR STREAMING TRAFFIC WITH ACTIVITY DETECTION FUNCTION

Kiril Kassev, Yakim Mihov, Boris Tsankov

http://www.foibg.com/ibs_isc/ibs-14/ibs-14-p06.pdf

Abstract: Subtitling is used extensively by broadcasters both for foreign-language subtitling and as an access service to help people with a disability access television programs. The European Broadcaster Union (EBU) has create a working group in 2009-01-21 for use the standard Distribution Format Exchange Profile 1 (DFXP) for exchange subtitling information in XML with the Material Exchange Format 2 (MXF) file format. At this moment, in order to help the deaf people, broadcaster produces textual data in a file with a time-code referred of the picture. Subtitle live events require speech recognition or special keyboards where the words are presented as a union of several keys pressed. In this paper, we present an initiative based in B2B standards for the exchange, production, and broadcast multilingual subtitle for live Opera production.

Keywords: MPEG2, Multimedia, DTV, Subtitling, Broadcasting.

Link:

MULTILANGUAGE OPERA SUBTITLING EXCHANGE BETWEEN PRODUCTION AND BROADCASTER COMPANIES

Jesús Martínez Barbero, Manuel Bollaín Pérez

http://www.foibg.com/ibs_isc/ibs-14/ibs-14-p05.pdf

Abstract: This paper presents a comparison of four representative discretization methods from different classes to be used with so called PGN-classifier which deals with categorical data. We examine which of them supplies more convenient discretization for PGN Classification Method. The experiments are provided on the base of UCI repository data sets. The comparison tests were provided using an experimental classification machine learning system "PaGaNe", which realizes Pyramidal Growing Network (PGN) Classification Algorithm. It is found that in general, PGN-classifier trained on data preprocessed by Chi-merge achieve lower classification error than those trained on data preprocessed by the other discretization methods. The comparison of PGN-classifier, trained with Chi-merge-discretizator with other classifiers (realized in WEKA system) shows good results in favor of PGNclassifier.

Keywords: Data Mining, Machine Learning, Discretization, Data Analysis, Pyramidal Growing Networks

Link:

COMPARISON OF DISCRETIZATION METHODS FOR PREPROCESSING DATA FOR PYRAMIDAL GROWING NETWORK CLASSIFICATION METHOD

Ilia Mitov, Krassimira Ivanova, Krassimir Markov, Vitalii Velychko, Peter Stanchev, Koen Vanhoof

http://www.foibg.com/ibs_isc/ibs-14/ibs-14-p04.pdf