Reusch Computational Intelligence, Theory and Applications

International Conference 9th Fuzzy Days in Dortmund, Germany, Sept. 18-20, 2006 Proceedings
1. Auflage 2006
ISBN: 978-3-540-34783-5
Verlag: Springer Berlin Heidelberg
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, Band 38, 802 Seiten

Reihe: Advances in Intelligent and Soft Computing

This book constitutes the refereed proceedings of the 9th Dortmund Fuzzy Days, Dortmund, Germany, 2006. This conference has established itself as an international forum for the discussion of new results in the field of Computational Intelligence. The papers presented here, all thoroughly reviewed, are devoted to foundational and practical issues in fuzzy systems, neural networks, evolutionary algorithms, and machine learning and thus cover the whole range of computational intelligence.

Autoren/Hrsg.

Reusch, Bernd

Inhaltsverzeichnis

1;Preface;6
1.1;Programme Chairs;7
1.2;Programme Committee;7
1.3;Local Organization;7
2;Contents;8
3;List of Contributors;18
4;Plenary Talk;31
4.1;From Search Engines to Question-Answering Systems: The Problems of World Knowledge, Relevance, Deduction, and Precisiation;31
5;Invited Session: Fuzzy Multiperson and Multicriteria Decisions Modelling;35
5.1;A Fuzzy Approach to Optimal R&D Project Portfolio Selection;35
5.1.1;1 Introduction;35
5.1.2;2 Real Options for R&D Portfolios;37
5.1.3;3 A Hybrid Approach to Real Option Valuation;39
5.1.4;4 A Possibilistic Approach to R&D Portfolio Selection;40
5.1.5;5 Summary;42
5.1.6;References;42
5.2;Choquet Integration and Correlation Matrices in Fuzzy Inference Systems;45
5.2.1;References;47
5.3;Linguistic Summarization of Some Static and Dynamic Features of Consensus Reaching;49
5.3.1;1 Introduction;49
5.3.2;2 Degrees of Consensus under Fuzzy Preferences and a Fuzzy Majority;50
5.3.3;3 A Consensus Reaching Process and Linguistic Data Summarization;52
5.3.4;4 A Dynamic View of Consensus Reaching;55
5.3.5;5 Concluding Remarks;56
5.3.6;References;56
5.4;Consistency for Nonadditive Measures: Analytical and Algebraic Methods;59
5.4.1;1 Choosing Among Several Alternatives;59
5.4.2;2 Fuzzy Measures on the Set of Objectives;60
5.4.3;3 Aggregation of Fuzzy Measures with Respect to a t- Conorm;62
5.4.4;4 Algebraic and Geometric Representations by Means of Hypergroups;64
5.4.5;5 Assessment of Consistent Measures to the Objectives;67
5.4.6;6 An Algorithm to Compare the Alternatives;69
5.4.7;7 An Extension to the Fuzzy Measures of Type 2;69
5.4.8;References;70
6;Neural Nets;71
6.1;Neuro-Fuzzy Kolmogorov’s Network with a Modified Perceptron Learning Rule for Classification Problems;71
6.1.1;1 Introduction;71
6.1.2;2 Network Architecture;72
6.1.3;3 Learning Algorithm;73
6.1.4;4 Experiments;76
6.1.5;5 Conclusion;78
6.1.6;References;79
6.2;A Self-Tuning Controller for Teleoperation System using Evolutionary Learning Algorithms in Neural Networks;81
6.2.1;1 Introduction;81
6.2.2;2 Neural Controller and Evolutionary Programming;84
6.2.3;3 Fitness Function;86
6.2.4;4 Simulation Results;87
6.2.5;5 Conclusion;89
6.2.6;References;89
6.3;A Neural-Based Method for Choosing Embedding Dimension in Chaotic Time Series Analysis;91
6.3.1;1 Introduction;91
6.3.2;2 Characteristics of Chaotic Systems;93
6.3.3;3 Maximal Lyapunov Exponent;93
6.3.4;4 Phase Space Reconstruction, Embedding Theorems;94
6.3.5;5 Choosing the Delay Time;95
6.3.6;6 Embedding Dimension Estimation;96
6.3.7;7 Predictive Method for Minimum Embedding Dimension Estimation;97
6.3.8;8 Indirect Method for Maximal Lyapunov Estimation;99
6.3.9;9 Simulation Results;99
6.3.10;10 Conclusion;102
6.3.11;References;103
6.4;On Classification of Some Hop.eld-Type Learning Rules via Stability Measures;105
6.4.1;1 Introduction;105
6.4.2;2 The Hopfield Model and Its Storage Capacity;106
6.4.3;3 Some Learning Methods and Architectures;106
6.4.4;4 Empirical Analysis of Storage Capacity;108
6.4.5;5 Classification of Hopfield Memories via a Stability Measure;109
6.4.6;6 Classification of the Models via Experimental Analysis;110
6.4.7;7 Conclusion;112
6.4.8;References;113
7;Applications I;115
7.1;A New Genetic Based Algorithm for Channel Assignment Problems;115
7.1.1;1 Introduction;115
7.1.2;2 Channel Assignment Problem Formulation;116
7.1.3;3 Conventional Genetic Algorithm;117
7.1.4;4 Our New Method Based on Genetic Algorithm;117
7.1.5;5 Simulation Results;119
7.1.6;6 Conclusion;120
7.1.7;Acknowledgment;121
7.1.8;References;121
7.2;Max-Product Fuzzy Relational Equations as Inference Engine for Prediction of Textile Yarn Properties;123
7.2.1;1 Introduction;123
7.2.2;2 Max-Prod Fuzzy Linear Equations as Inference Engine;124
7.2.3;3 Implementation Methodology;127
7.2.4;4 Numerical Example;130
7.2.5;5 Discussion;131
7.2.6;6 Conclusions;132
7.2.7;Acknowledgments;132
7.2.8;References;132
7.3;Automatic Defects Classification and Feature Extraction Optimization;135
7.3.1;1 Motivation;135
7.3.2;2 Automatic Classification System;136
7.3.3;3 Feature Extraction Technologies;136
7.3.4;4 Classification Results;140
7.3.5;5 Optimization of Feature Extraction;140
7.3.6;6 Summary;145
7.3.7;References;145
7.4;Short-Term Load Forecasting in Power System Using Least Squares Support Vector Machine;147
7.4.1;1 Introduction;147
7.4.2;2 Review of LS-SVM;148
7.4.3;3 Short-Term Forecasting Using LS-SVM;150
7.4.4;4 Examples;152
7.4.5;5 Conclusions;154
7.4.6;6 Copyright Form;154
7.4.7;References;155
7.4.8;Appendix: Springer-Author Discount;156
8;Plenary Talk;157
8.1;Fifteen Years of Fuzzy Logic in Dortmund;157
8.1.1;1 Introduction;157
8.1.2;2 Basic Operations;158
8.1.3;3 Fuzzy Logic;159
8.1.4;4 Inference Methods;160
8.1.5;5 Fuzzy Relations;161
8.1.6;6 Fuzzy Sets, Related Concepts, and Applications;161
8.1.7;7 Concluding Remarks;163
8.1.8;References;165
9;Invited Session: Intuitionistic Fuzzy Sets and Generalized Nets I;169
9.1;Intuitionistic Fuzzy Graphs;169
9.1.1;1 Introduction;169
9.1.2;2 Preliminaries;170
9.1.3;3 Properties;173
9.1.4;4 Conclusion;179
9.1.5;Acknowledgement;179
9.1.6;References;179
9.2;On Some Intuitionistic Properties of Intuitionistic Fuzzy Implications and Negations;181
9.2.1;1 Introduction: on Some Previous Results;181
9.2.2;2 Main Results;182
9.2.3;3 Conclusion;186
9.2.4;References;188
9.3;On Intuitionistic Fuzzy Negations;189
9.3.1;1 Introduction: on Some Previous Results;189
9.3.2;2 Main Results;191
9.3.3;3 Conclusion: a New Argument that the Intuitionistic Fuzzy Sets Have Intuitionistic Nature;195
9.3.4;References;197
10;Invited Session: Soft Computing Techniques for Reputation and Trust I;199
10.1;A Simulation Model for Trust and Reputation System Evaluation in a P2P Network;199
10.1.1;1 Model Requirements;200
10.1.2;2 The Peer Behavior Model;205
10.1.3;3 The Simulation Model;206
10.1.4;4 Conclusions;209
10.1.5;References;209
10.2;A Fuzzy Trust Model Proposal to Ensure the Identity of a User in Time;211
10.2.1;1 Introduction;211
10.2.2;2 User Authentication Systems and Their Trustfulness;212
10.2.3;3 Architecture of the Model;216
10.2.4;4 Trust Model Rules;218
10.2.5;5 Conclusions;219
10.2.6;Acknowledgments;220
10.2.7;References;220
10.3;Quantification of the Effectiveness of the Markov Model for Trustworthiness Prediction;221
10.3.1;1 Introduction;221
10.3.2;2 Exogenous Parameters Used in the Simulation;223
10.3.3;3 Behaviors of the Agents in the Prototype Simulation;223
10.3.4;4 Effectiveness of Markov Model on Trust Based Decision Making;225
10.3.5;5 Conclusions and Future Work;229
10.3.6;References;230
11;Applications II;231
11.1;Fuzzy-Genetic Methodology for Web-based Computed- Aided Diagnosis in Medical Applications;231
11.1.1;1 Introduction;231
11.1.2;2 Materials and Methods;233
11.1.3;3 Case Study;239
11.1.4;4 Concluding Remarks;242
11.1.5;5 Acknowledgments;242
11.1.6;References;242
11.2;Weight Optimization for Loan Risk Estimation with Genetic Algorithm;245
11.2.1;1 Introduction;245
11.2.2;2 GA in Weight Optimization Task;246
11.2.3;3 An Individual Fitness;247
11.2.4;4 Crossover;248
11.2.5;5 Mutation;248
11.2.6;6 Experimental Results;249
11.2.7;7 Conclusions;250
11.2.8;Acknowledgments;250
11.2.9;References;250
11.3;A Fuzzy Feature Extractor Neural Network and its Application in License Plate Recognition;253
11.3.1;1 Introduction;253
11.3.2;2 Feature Extractor Fuzzy Neural Network;254
11.3.3;3 License Plate Type Recognition;255
11.3.4;4 Results;256
11.3.5;References;257
12;Invited Session: Intuitionistic Fuzzy Sets and Generalized Nets II;259
12.1;Nearest Interval Approximation of an Intuitionistic Fuzzy Number;259
12.1.1;1 Intuitionistic Fuzzy Numbers;259
12.1.2;2 Distances between Intuitionistic Fuzzy Numbers;261
12.1.3;3 Nearest Interval Approximation of Intuitionistic Fuzzy Numbers;262
12.1.4;4 Properties of the Nearest Intervals Approximation of Intuitionistic Fuzzy Numbers;268
12.1.5;References;270
12.2;On Intuitionistic Fuzzy Expert Systems With Temporal Parameters;271
12.2.1;1 Introduction;271
12.2.2;2 Short Remarks on IFL;271
12.2.3;3 Main Results;275
12.2.4;4 Conclusion;279
12.2.5;References;279
12.3;Generalized Fuzzy Cardinalities of IF Sets;281
12.3.1;1 Introduction;281
12.3.2;2 Preliminaries;281
12.3.3;3 Fuzzy Cardinality of Fuzzy Sets;285
12.3.4;4 Fuzzy Cardinality of IF Sets;287
12.3.5;Conclusion;290
12.3.6;References;291
13;Invited Session: Soft Computing Techniques for Reputation and Trust II;293
13.1;Towards Usage Policies for Fuzzy Inference Methodologies for Trust and QoS Assessment;293
13.1.1;1 Related Work;294
13.1.2;2 Fuzzy Trust and QoS Assessment;295
13.1.3;3 Fuzzy Rules and Inference Methodologies;297
13.1.4;4 Usage Policies for Fuzzy Inference Methodologies;299
13.1.5;5 Conclusion;302
13.1.6;References;303
13.2;Simulating a Trust-Based Peer-to-Peer Metadata Publication Center;305
13.2.1;1 Introduction;305
13.2.2;2 The Trust Layer Architecture;306
13.2.3;3 Setting a Trust Layer Simulator;308
13.2.4;4 Some Examples of Simulations;309
13.2.5;Acknowledgments;309
13.2.6;References;310
13.3;The Complex Facets of Reputation and Trust;311
13.3.1;1 Introduction;311
13.3.2;2 Fundamentals;312
13.3.3;3 State-of-the-Art;312
13.3.4;4 Propagation of Reputation Information;315
13.3.5;5 Bounded Rationality;317
13.3.6;6 Behavioral Evolution;318
13.3.7;7 Second-Order Defection Problem;319
13.3.8;8 Inhomogeneous Interactions;320
13.3.9;9 Identity Stability;321
13.3.10;10 Conclusions;322
13.3.11;References;323
14;Theory I;325
14.1;Fuzzy Covering Relation and Ordering: An Abstract Approach;325
14.1.1;1 Covering and Order in Crisp Case;325
14.1.2;2 Lattice-Valued Fuzzy Order and Covering;327
14.1.3;3 Examples;329
14.1.4;4 Conclusion;330
14.1.5;Acknowledgment;330
14.1.6;References;330
14.2;Measures of Differentiability;331
14.2.1;1 Introduction and Preliminaries;331
14.2.2;2 Measures of Differentiability;333
14.2.3;Acknowledgment;336
14.2.4;References;336
14.3;Lipschitz Continuity of Triangular Norms;339
14.3.1;1 Introduction;339
14.3.2;2 k-Lipschitz t-Norms;341
14.3.3;3 Boundaries of the Class of k-lp-Lipschitz t-Norms;343
14.3.4;4 Transformations of k-Lipschitz t-Norms;349
14.3.5;Acknowledgment;351
14.3.6;References;351
15;Plenary Talk;353
15.1;Formal Models of Knowledge Operators: Rough- Set- Style and Fuzzy- Set- Style Approaches;353
15.1.1;References;353
16;Invited Session: Looking at Language with Fuzzy Logic;355
16.1;Using a Fuzzy Model for Combining Search Results from Di . erent Information Sources to Build a Metasearch Engine;355
16.1.1;1 Introduction;355
16.1.2;2 Foundations;356
16.1.3;3 Metasearch;358
16.1.4;4 Design and Implementation;359
16.1.5;5 Conclusion;362
16.1.6;Acknowledgment;364
16.1.7;References;364
16.2;Some Fuzzy Counterparts of the Language uses of And and Or;365
16.2.1;1 Introduction;365
16.2.2;2 Standard Models of And (Or) in Fuzzy Logic;368
16.2.3;3 Relations Induced by an Operation;370
16.2.4;4 Conjunction and Weak Conjunction;370
16.2.5;5 Inclusive or: Disjunction and Weak Disjunction;374
16.2.6;6 Exclusive or: Symmetric Di.erence;377
16.2.7;7 Conclusions;380
16.2.8;Acknowledgment;381
16.2.9;References;381
16.3;Fuzzy Sets Versus Language;383
16.3.1;1 Introduction;383
16.3.2;2 On General Theories of Fuzzy Sets;385
16.3.3;3 Decomposable Theories of Fuzzy Sets;386
16.3.4;4 Examples Suggesting a Generalization of the Theories ( T, S, N);390
16.3.5;5 Antonyms;393
16.3.6;6 Last Comments;395
16.3.7;Acknowledgment;396
16.3.8;References;396
17;Theory II;397
17.1;Some Properties of Fuzzy Languages;397
17.1.1;1 Introduction;397
17.1.2;2 Fuzzy Chomsky Languages;398
17.1.3;3 Fuzzy Petri Net Languages;400
17.1.4;4 Fuzzy Lindenmayer Languages;402
17.1.5;5 Closing Remarks;403
17.1.6;References;403
17.2;General Form of Lattice Valued Intuitionistic Fuzzy Sets;405
17.2.1;1 Introduction;405
17.2.2;2 Results;407
17.2.3;Acknowledgement;410
17.2.4;References;411
17.3;A Note on Generated Pseudo-Operations with Two Parameters as a base for the Generalized Pseudo- Laplace Type Transform;413
17.3.1;1 Introduction;413
17.3.2;2 Preliminary Notions;414
17.3.3;3 The Generalized (,)-Laplace Transform Based onGenerated Pseudo-Operations with Two Parameters;416
17.3.4;4 Pseudo-Aggregation Operators Based on theGeneralized (,)-Laplace Transforms;421
17.3.5;5 Conclusion;423
17.3.6;Acknowledgments;423
17.3.7;References;423
18;Theory III;425
18.1;Fuzzy All-Pairs Shortest Paths Problem;425
18.1.1;1 Introduction;425
18.1.2;2 Crisp Problem and Its Time Complexity;426
18.1.3;3 Fuzzy Version of the APSPP;431
18.1.4;4 Conclusions;433
18.1.5;Acknowledgement;434
18.1.6;References;434
18.2;Optimal Toll Charges in a Fuzzy Flow Problem;435
18.2.1;1 Introduction;435
18.2.2;2 The Fuzzy Flow Problem;436
18.2.3;3 The Toll Finding Problem and Its Reformulation;437
18.2.4;4 Solution Algorithm for the Bilevel Programming Problem;439
18.2.5;5 Conclusion;442
18.2.6;References;442
18.3;Modified Interval Global Weights in AHP;445
18.3.1;1 Introduction;445
18.3.2;2 Interval Priority Weights from Crisp Pairwise Comparisons;446
18.3.3;3 Numerical Example;451
18.3.4;4 Conclusion;453
18.3.5;References;454
19;Plenary Talk;455
19.1;Fuzzy Approaches to Trust Management;455
19.1.1;1 Introduction;455
19.1.2;2 Trust Management Systems;456
19.1.3;3 Trust Negotiation Protocols;457
19.1.4;4 Reputation-Based Protocols;458
19.1.5;5 Aggregation-Based Methods for Computing Trust;459
19.1.6;6 Fuzzy Rule-Based Methods for Trust Management;463
19.1.7;7 Conclusions;463
19.1.8;Acknowledgments;464
19.1.9;References;464
20;Invited Session: Complex-Valued Neural Networks;467
20.1;Proposal of Holographic 3D-Movie Generation Using Coherent Neural- Network Interpolation;467
20.1.1;1 Introduction;467
20.1.2;2 Hologram Interpolation Utilizing Generalization;467
20.1.3;3 Simulation Experiment;468
20.1.4;4 Summary;468
20.1.5;References;468
20.2;Blur Identification Using Neural Network for Image Restoration;471
20.2.1;1 Introduction;471
20.2.2;2 Image Restoration Problem;473
20.2.3;3 Multilayer Neural Network Based On Multi-Valued Neurons;474
20.2.4;4 Simulations;478
20.2.5;5 Conclusions;483
20.2.6;Acknowledgment;483
20.2.7;References;483
20.3;Solving the Parity n Problem and Other Nonlinearly Separable Problems Using a Single Universal Binary Neuron;487
20.3.1;1 Introduction;487
20.3.2;2 UBN and MVN;488
20.3.3;3 Solving the Parity n Problem Using a Single UBN;495
20.3.4;4 Implementation of the Edge Detecting Boolean Functions;499
20.3.5;4 Conclusions;500
20.3.6;References;500
20.4;Some Novel Real/Complex-Valued Neural Network Models;503
20.4.1;1 Introduction;503
20.4.2;2 Continuous Time Perceptron and Generalizations;504
20.4.3;3 A New Mathematical Model of Neuron/Single Perceptron;505
20.4.4;4 Abstract Mathematical Structure of Neuronal Models;506
20.4.5;5 Finite Impulse Response Model of Synapses: Neural Networks;508
20.4.6;6 Novel Continuous Time Associative Memory;509
20.4.7;7 Multi-Dimensional Generalizations;511
20.4.8;8 Generalization to Complex-Valued Neural Networks;512
20.4.9;9 Conclusions;512
20.4.10;References;513
21;Theory IV;515
21.1;Extending the Fuzzy Rule Interpolation “ FIVE” by Fuzzy Observation;515
21.1.1;1 Introduction;516
21.1.2;2 The concept of Vague Environment;517
21.1.3;3 Approximate Scaling Function;519
21.1.4;4 Shepard Interpolation for Fuzzy Reasoning: “FIVE”;520
21.1.5;5 Fuzzy Observation by Merging Vague Environments;522
21.1.6;6 Example;524
21.1.7;7 Conclusions;525
21.1.8;Acknowledgement;526
21.1.9;References;526
21.2;Fuzzy Rule Interpolation Based on Polar Cuts;529
21.2.1;1 A Brief Overview of Fuzzy Rule Interpolation Methods;530
21.2.2;2 The Structure of the Proposed Method;530
21.2.3;3 Fuzzy Set Interpolation Based on Linguistic Term Shifting and Polar Cuts;531
21.2.4;4 The Position of the Consequent Sets;534
21.2.5;5 Single Rule Reasoning Based on Polar Cuts;535
21.2.6;6 Numerical Examples;538
21.2.7;7 Conclusions;539
21.2.8;8 Acknowledgments;539
21.2.9;References;540
21.3;Approximate Reasoning Using Fodor’s Implication;543
21.3.1;1 Introduction;543
21.3.2;2 Basic Concepts;544
21.3.3;3 Generalized Modus Ponens with Fodor’s Implication;545
21.3.4;References;549
22;Plenary Talk;551
22.1;Brain-, Gene-, and Quantum-Inspired Computational Intelligence: Challenges and Opportunities;551
22.1.1;1 Introduction: Brain, Gene, and Quantum Levels of Information Processing in the Brain as Inspirations for ANN and CI Models;552
22.1.2;2 Some Brain-Inspired ECOS Models;553
22.1.3;3 Brain–Gene-Inspired CNGM;559
22.1.4;4 Quantum-Inspired Evolving Connectionist Models;563
22.1.5;5 Conclusions and Directions for Further Research;569
22.1.6;Acknowledgement;569
22.1.7;References;570
23;Invited Session: Intelligent Data Mining;575
23.1;Incremental Learning for E-mail Classification;575
23.1.1;1 Introduction;575
23.1.2;2 Incremental Learning;576
23.1.3;3 Partial Memory Incremental Learning Algorithm FLORA2;577
23.1.4;4 Experiments;580
23.1.5;5 Conclusions;582
23.1.6;References;583
23.2;Reduction of Search Space for Instance-Based Classifier Combination;585
23.2.1;1 Introduction;585
23.2.2;2 Reduction of Search Space via Data Condensation;586
23.2.3;3 Adaptive Classi.er Combination;587
23.2.4;4 Case Study: Credit Scoring Data Set;588
23.2.5;5 Conclusions;589
23.2.6;Acknowledgement;590
23.2.7;References;590
24;Invited Session: Preferences and Decisions;591
24.1;Linguistic Matrix Aggregation Operators: Extensions of the Borda Rule;591
24.1.1;1 Introduction;591
24.1.2;2 Extending the Classic Borda Rule to a Linguistic Framework;593
24.1.3;3 Linguistic Matrix Aggregation Operators and Decision Rules;596
24.1.4;4 Social Choice Type Properties;600
24.1.5;5 Some Further Extensions and Concluding Remarks;602
24.1.6;Acknowledgments;603
24.1.7;References;603
25;Evolutionary Algorithms;607
25.1;An Evolutionary Algorithm for the Biobjective QAP;607
25.1.1;1 Introduction;607
25.1.2;2 The EC-Memory Method;608
25.1.3;3 The New Algorithm;610
25.1.4;4 Experimental Results;612
25.1.5;5 Summary;615
25.1.6;Acknowledgements;615
25.1.7;References;615
25.2;On a Hill-Climbing Algorithm with Adaptive Step Size: Towards a Control Parameter- Less Black- Box Optimisation Algorithm;617
25.2.1;1 Introduction;617
25.2.2;2 Self-Adaptive Step-size Search (SASS);618
25.2.3;3 Experiments;618
25.2.4;4 Experimental Results;620
25.2.5;5 Analysis of the SASS Algorithm;620
25.2.6;6 Conclusions and Future Work;624
25.2.7;References;625
25.3;Self-Adaptive Baldwinian Search in Hybrid Genetic Algorithms;627
25.3.1;1 Introduction;627
25.3.2;2 Evolutionary Self-Adaptation and Duration of Local Search;628
25.3.3;3 Experiments;629
25.3.4;4 Discussion;630
25.3.5;5 Conclusions;631
25.3.6;References;631
25.4;Intragenerational Mutation Shape Adaptation;633
25.4.1;1 Introduction;633
25.4.2;2 The DCMA-ES Algorithm;634
25.4.3;3 Experimental Results;638
25.4.4;4 Conclusions;642
25.4.5;References;642
26;Theory V;645
26.1;The Choquet-Integral as an Aggregation Operator in Case- Based Learning;645
26.1.1;1 Introduction;645
26.1.2;2 Nearest Neighbor Estimation;646
26.1.3;3 The Cho-k-NN Method;647
26.1.4;4 Empirical Validation;651
26.1.5;5 Speci.cation of Similarity Measures;653
26.1.6;6 Concluding Remarks;655
26.1.7;References;656
26.2;Fuzzy Sets and Multicriteria Decision Making;659
26.2.1;1 Introduction;659
26.2.2;2 Fuzzy Measures/Aggregation Approach;660
26.2.3;3 Fuzzy Logic Based Construction of Preference Relations;661
26.2.4;4 Preference Relations Based on Orderings of Fuzzy Quantities;662
26.2.5;5 Conclusion;663
26.2.6;Acknowledgment;664
26.2.7;References;664
26.3;Fuzzy Reinforcement Learning for Routing in Wireless Sensor Networks;667
26.3.1;1 Introduction;667
26.3.2;2 Assumptions and Problem Formulation;668
26.3.3;3 Fuzzy Reinforcement Learning for Routing in Wireless Sensor Networks;669
26.3.4;4 Experimental Results;673
26.3.5;5 Conclusion;673
26.3.6;References;674
26.4;Outlier Resistant Recursive Fuzzy Clustering Algorithms;677
26.4.1;1 Introduction;677
26.4.2;2 Recursive Fuzzy Clustering Algorithm;678
26.4.3;3 Experiments;679
26.4.4;4 Conclusion;681
26.4.5;References;681
27;Invited Session: Fuzzy Sets – 40 years after;683
27.1;Fuzzy Set Theory – 40 Years of Foundational Discussions;683
27.1.1;1 Introduction;683
27.1.2;2 Model Oriented Constructions;684
27.1.3;3 Axiomatizations;688
27.1.4;4 Category Theoretic Approaches;689
27.1.5;References;693
27.2;Fuzzy Control – Expectations, Current State, and Perspectives;697
27.2.1;1 Success of Fuzzy Control;698
27.2.2;2 General Problems;701
27.2.3;3 Specific Problems;702
27.2.4;4 Conclusions and Perspectives;704
27.2.5;Acknowledgments;704
27.2.6;References;704
27.3;Fuzzy Sets in Categories of Sets with Similarity Relations;707
27.3.1;1 Introduction;707
27.3.2;2 Fuzzy Sets in -sets;708
27.3.3;3 Properties of functors;711
27.3.4;References;712
27.4;Fuzzy Sets as a Special Mathematical Model of Vagueness Phenomenon;713
27.4.1;1 Introduction;713
27.4.2;2 Uncertainty and Vagueness;714
27.4.3;3 Actuality and Potentiality;715
27.4.4;4 Fuzzy Sets Naturally Emerge as a Graded Model of Vagueness;716
27.4.5;5 Future Development of Fuzzy Set Theory;718
27.4.6;6 Conclusion;719
27.4.7;References;719
27.5;Fuzzy IF-THEN Rules from Logical Point of View;721
27.5.1;1 Introduction;721
27.5.2;2 Special Theory of Fuzzy IF-THEN Rules;724
27.5.3;3 Conclusion;727
27.5.4;References;727
28;Applications III;683
28.1;Synthesizing Adaptive Navigational Robot Behaviours Using a Hybrid Fuzzy A* Approach;729
28.1.1;1 Introduction;729
28.1.2;2 General System Architecture (Fig. 1);730
28.1.3;3 The Evasion Algorithm;731
28.1.4;4 Cascade of Fuzzy Systems;736
28.1.5;5 Conclusions;739
28.1.6;References;739
28.2;Fuzzy Impulse Noise Reduction Methods for Color Images;741
28.2.1;1 Introduction;741
28.2.2;2 Filters for Noise Reduction;742
28.2.3;3 Comparative Study;744
28.2.4;4 Conclusion;749
28.2.5;References;749
28.3;Use of Variable Fuzzy Sets Methods for Desertification Evaluation;751
28.3.1;1 Introduction;751
28.3.2;2 Principle of VFS;752
28.3.3;3 VFS for Comprehensive Evaluation of the Deserti . cation Degree;754
28.3.4;4 Conclusion;759
28.3.5;Acknowledgments;760
28.3.6;References;760
28.4;A Fuzzy Ultrasonic System for Estimating Degradation of Insulating Oil;763
28.4.1;1 Introduction;763
28.4.2;2 Preliminaries;764
28.4.3;3 Fuzzy Ultrasonic Estimation System;765
28.4.4;4 Experimental Results;769
28.4.5;5 Conclusions;769
28.4.6;References;770
28.5;A Genetic Algorithm-Based Fuzzy Inference System in Prediction of Wave Parameters;771
28.5.1;1 Introduction;771
28.5.2;2 Fuzzy Inference Systems (FISs);772
28.5.3;3 Subtractive Clustering;774
28.5.4;4 Hybrid GA-ANFIS Model;776
28.5.5;5 Application;777
28.5.6;6 Summary and Conclusions;779
28.5.7;References;779
29;Poster Contributions;781
29.1;Estimation of Degree of Polymerisation and Residual Age of Transformers Based on Furfural Levels in Insulating Oil Through Generalized Regression Neural Networks;781
29.1.1;1 Introduction;781
29.1.2;2 Transformer Insulation Measurement and Residual Life Assessment;782
29.1.3;3 Choice of Arti.cial Neural Networks;782
29.1.4;4 Application of Generalized Regression Neural Networks;783
29.1.5;5 Conclusion;785
29.1.6;Acknowledgement;786
29.1.7;References;786
29.2;Fuzzy Shortest Paths in Fuzzy Graphs;787
29.2.1;1 Introduction;787
29.2.2;2 Preliminaries;788
29.2.3;3 The Proposed Method;789
29.2.4;4 The Proposed Algorithm;792
29.2.5;5 Simulation;793
29.2.6;6 Conclusion;794
29.2.7;References;794
29.3;Improving Vegas Algorithm Using PID and Fuzzy PID Controllers;795
29.3.1;1 Introduction;795
29.3.2;2 Congestion Control;795
29.3.3;3 TCP Vegas;796
29.3.4;4 Developing TCP Vegas using PID Controller;797
29.3.5;5 Developing TCP Vegas using Fuzzy PID Controller;797
29.3.6;6 Performance Evaluation;800
29.3.7;7 Conclusion;804
29.3.8;8 Acknowledgment;805
29.3.9;References;805
29.4;A Fuzzy-Based Automation Level Analysis in Irrigation Equipment;807
29.4.1;1 Introduction;807
29.4.2;2 Automation Level and Automation Threshold;808
29.4.3;3 Automation Level Determination;808
29.4.4;4 Energy Resources;809
29.4.5;5 Field;810
29.4.6;6 Mechanization;813
29.4.7;7 Fuzzy Analyzer;813
29.4.8;8 Simulation Results;815
29.4.9;9 Conclusion;816
29.4.10;References;817
29.5;Motorized Skateboard Stabilization Using Fuzzy Controller;819
29.5.1;1 Introduction;819
29.5.2;2 Skateboard Model;820
29.5.3;3 Body Model;821
29.5.4;4 Signal processor;822
29.5.5;5 Fuzzy Tuner;823
29.5.6;6 Simulation Results;825
29.5.7;7 Conclusion;827
29.5.8;References;828
30;Index;831

Leseproben

Fuzzy Control – Expectations, Current State, and Perspectives (p. 667)

Mirko Navara and Milan Petr´ýk

Summary.

We summarize the history of fuzzy sets. We try to find the reasons why fuzzy control has been so successful in applications. This is mainly explained by the fact that fuzzy logic created an alternative to exact computation and it better fits to the human way of reasoning.

We point out some aspects in which current fuzzy systems are not completely satisfactory and directions in which they should develop in the future.

Key words:

Fuzzy set, Fuzzy control, Computational complexity, Fuzzy arithmetic, Stability.

The idea of partial truth and partial membership is old and it has been rediscovered many times (e.g., (4, 7, 13)). However, the seminal paper (28) has opened a new epoch of its rapid development. Our first question is why exactly this work initiated a revolution if many theoretical results (see (4, 24)) have been derived before and remained almost unnoticed.

One reason is that Zadeh expressed this idea in a way accepted by experts in many fields, not only theoretical, but also applied, even by engineers. The preceding papers were recognized only by a limited community of mathematicians. Now the principle was expressed in a way understandable to everybody and in a context drawing new horizons and capabilities of the new technology based on it. It might have been crucial that the applications in control theory followed very soon (14, 26, 29).

Their success ensures permanent interest of industrial partners and financial support of this field. The second reason of success of fuzzy logic in Zadeh’s approach is the state of control theory in the sixties. Preceding development of computers and cybernetics has brought ambitious expectations which have been satisfied only partially. The rapid development of control theory, as initiated by Wiener, has slowed down.

It solved successfully some problems, in particular in control of linear systems, but it has encountered di.culties in control of systems with high non-linearity. These were partially solved by the developing non-linear control theory and by adaptive control, but this efort has brought much more complex questions without a clear trend to their satisfactory solutions. We bring arguments that in some sense the same happened to fuzzy control a few decades later.

The third reason is a disillusion from the limits of computational power. At the first moment, people were fascinated by the newly open possibility of cheap high-precision computations ofered by computers. However, they recognized soon that some solutions are far from satisfactory. Simplified models failed to describe important features of real systems and the solutions did not perform well on some real-world systems.

Then it was found out that supreme precision is not as important. Instead of that, we need to describe (at least roughly) the complexity of the surrounding world. This requires a representation of numerous relations which are not precisely known, but whose effect is at least intuitively understood by humans. Fuzzy logic offered a tool allowing to implement these ideas easily.