Fuzzy and Neural Approaches in Engineering

Neural networks and fuzzy systems represent two distinct technologies that deal with uncertainty. This definitive book presents the fundamentals of both technologies, and demonstrates how to combine the unique capabilities of these two technologies for the greatest advantage. Steering clear of unnecessary mathematics, the book highlights a wide range of dynamic possibilities and offers numerous examples to illuminate key concepts. It also explores the value of relating genetic algorithms and expert systems to fuzzy and neural technologies.

LEFTERI H. TSOUKALAS, PhD, is on the faculty of the School of Nuclear Engineering at Purdue University and is an active industrial consultant and speaker. ROBERT E. UHRIG, PhD, holds a joint appointment as Distinguished Professor in the Nuclear Engineering Department at the University of Tennessee and Distinguished Scientist in the Instrumentation and Control Division at the Oak Ridge National Laboratory. He is the author of Random Noise Techniques in Nuclear Reactor Systems.

Foreword

xiii

Preface

xvii

Introduction to Hybrid Artificial Intelligence Systems

1

(10)

Introduction

1

(1)

Neural Networks and Fuzzy Logic Systems

2

(1)

The Progress in Soft Computing

3

(2)

Intelligent Management of Large Complex Systems

5

(2)

Structure of this Book

7

(1)

Problems and Programs Available on the Internet

8

(3)

References

9

(2)

I FUZZY SYSTEMS: CONCEPTS AND FUNDAMENTALS

11

(178)

Foundations of Fuzzy Approaches

13

(36)

From Crisp to Fuzzy Sets

13

(2)

Fuzzy Sets

15

(2)

Basic Terms and Operations

17

(11)

Properties of Fuzzy Sets

28

(2)

The Extension Principle

30

(4)

Alpha-Cuts

34

(3)

The Resolution Principle

37

(1)

Possibility Theory and Fuzzy Probabilities

38

(11)

References

45

(1)

Problems

46

(3)

Fuzzy Relations

49

(28)

Introduction

49

(3)

Fuzzy Relations

52

(5)

Properties of Relations

57

(3)

Basic Operations with Fuzzy Relations

60

(5)

Composition of Fuzzy Relations

65

(12)

References

74

(1)

Problems

75

(2)

Fuzzy Numbers

77

(28)

Introduction

77

(2)

Representing Fuzzy Numbers

79

(5)

Addition

84

(6)

Subtraction

90

(5)

Multiplication

95

(4)

Division

99

(2)

Minimum and Maximum

101

(4)

References

102

(1)

Problems

103

(2)

Linguistic Descriptions and Their Analytical Forms

105

(40)

Fuzzy Linguistic Descriptions

105

(8)

Linguistic Variables and Values

113

(7)

Implication Relations

120

(5)

Fuzzy Inference and Composition

125

(11)

Fuzzy Algorithms

136

(9)

References

141

(1)

Problems

142

(3)

Fuzzy Control

145

(44)

Introduction

145

(6)

Fuzzy Linguistic Controllers

151

(12)

Defuzzification Methods

163

(13)

Issues Involved in Designing Fuzzy Controllers

176

(13)

References

185

(2)

Problems

187

(2)

II NEURAL NETWORKS: CONCEPTS AND FUNDAMENTALS

189

(218)

Fundamentals of Neural Networks

191

(38)

Introduction

191

(1)

Biological Basis of Neural Networks

192

(1)

Artificial Neurons

193

(3)

Artificial Neural Networks

196

(7)

Learning and Recall

203

(8)

Features of Artificial Neural Networks

211

(2)

Historical Development of Neural Networks

213

(8)

Separation of Nonlinearly Separable Variables

221

(8)

References

227

(1)

Problems

227

(2)

Backpropagation and Related Training Algorithms

229

(60)

Backpropagation Training

229

(5)

Widrow-Hoff Delta Learning Rule

234

(4)

Backpropagation Training for a Multilayer Neural Network

238

(10)

Factors That Influence Backpropagation Training

248

(7)

Sensitivity Analysis in a Backpropagation Neural Network

255

(2)

Autoassociative Neural Networks

257

(9)

An Alternate Approach to Neural Network Training

266

(4)

Modular Neural Networks

270

(4)

Recirculation Neural Networks

274

(5)

Functional Links

279

(1)

Cascade-Correlation Neural Networks

280

(1)

Recurrent Neural Networks

281

(8)

References

285

(2)

Problems

287

(2)

Competitive, Associative, and Other Special Neural Networks

289

(44)

Hebbian Learning

289

(1)

Cohen-Grossberg Learning

290

(6)

Associative Memories

296

(10)

Competitive Learning: Kohonen Self-Organizing Systems

306

(9)

Counterpropagation Networks

315

(4)

Probabilistic Neural Networks

319

(6)

Radial Basis Function Network

325

(1)

Generalized Regression Neural Network

326

(2)

Adaptive Resonance Theory (ART-1) Neural Networks

328

(5)

References

331

(1)

Problems

332

(1)

Dynamic Systems and Neural Control

333

(52)

Introduction

333

(1)

Linear Systems Theory

333

(8)

Adaptive Signal Processing

341

(4)

Adaptive Processors and Neural Networks

345

(8)

Neural Network Control

353

(10)

System Identification

363

(5)

Implementation of Neural Control Systems

368

(6)

Applications of Neural Networks in Noise Analysis

374

(6)

Time-Series Prediction

380

(5)

References

382

(1)

Problems

383

(2)

Practical Aspects of Using Neural Networks

385

(22)

Selection of Neural Networks for Solution to a Problem

385

(1)

Design of the Neural Network

386

(9)

Data Sources and Processing for Neural Networks

395

Data Representation

391

(4)

Scaling, Normalization, and the Absolute Magnitude of Data

395

(4)

Data Selection for Training and Testing

399

(2)

Training Neural Networks

401

(6)

References

405

(2)

III INTEGRATED NEURAL-FUZZY TECHNOLOGY

407

(114)

Fuzzy Methods in Neural Networks

409

(36)

Introduction

409

(1)

From Crisp to Fuzzy Neurons

410

(4)

Generalized Fuzzy Neuron and Networks

414

(2)

Aggregation and Activation Functions in Fuzzy Neurons

416

(2)

AND and OR Fuzzy Neurons

418

(3)

Multilayer Fuzzy Neural Networks

421

(2)

Learning and Adaptation in Fuzzy Neural Networks

423

(8)

Fuzzy ARTMAP

431

(3)

Fuzzy-Neural Hybrid Data Representation

434

(3)

Survey of Engineering Applications

437

(8)

References

440

(2)

Problems

442

(3)

Neural Methods in Fuzzy Systems

445

(26)

Introducing the Synergism

445

(2)

Fuzzy-Neural Hybrids

447

(3)

Neural Networks for Determining Membership Functions

450

(5)

Neural-Network-Driven Fuzzy Reasoning

455

(6)

Learning and Adaptation in Fuzzy Systems via Neural Methods

461

(5)

Adaptive Network-Based Fuzzy Inference Systems

466

(5)

References

468

(2)

Problems

470

(1)

Selected Hybrid Neurofuzzy Applications

471

(22)

Introduction

471

(1)

Neurofuzzy Interpolation

472

(2)

General Neurofuzzy Methodological Developments

474

(2)

Engineering Applications

476

(1)

Diagnostics in Complex Systems

477

(1)

Neurofuzzy Control Systems

478

(3)

Neurofuzzy Control in Robotics

481

(1)

Pattern Recognition and Image Enhancement

482

(1)

Medical and Environmental Imaging Using Neurofuzzy Methodologies

483

(1)

Transportation Control

484

(1)

Adaptive Fuzzy Systems

485

(1)

Inspection Using Neurofuzzy Methods

486

(1)

Neurofuzzy Methods in Financial Engineering

486

(1)

Commercial Neurofuzzy System Software

487

(6)

References

488

(5)

Dynamic Hybrid Neurofuzzy Systems

493

(28)

Introduction

493

(2)

Fuzzy-Neural Diagnosis for Vibration Monitoring

495

(5)

Decision Fusion by Fuzzy Set Operations

500

(4)

Hybrid Neurofuzzy Methodology for Virtual Measurements

504

(6)

Neurofuzzy Approaches to Anticipatory Control

510

(11)

References

516

(5)

IV OTHER ARTIFICIAL INTELLIGENCE SYSTEMS

521

(46)

Expert Systems in Neurofuzzy Systems

523

(16)

Introduction

523

(1)

Characteristics of Expert Systems

524

(1)

Components of an Expert System

525

(2)

Knowledge Representation and Inference

527

(2)

Uncertainty Management

529

(2)

State of the Art of Expert Systems

531

(1)

Use of Expert Systems

532

(2)

Expert Systems Used with Neural Networks and Fuzzy Systems

534

(1)

Potential Implementation Issues for Expert Systems

535

(4)

References

537

(1)

Problems

538

(1)

Genetic Algorithms

539

(22)

Introduction

539

(1)

Basic Concepts of Genetic Algorithms

540

(2)

Binary and Real-Value Representations of Chromosomes

542

(2)

Implementation of Genetic Algorithm Optimization

544

(2)

Fitness Functions

546

(6)

Application of Genetic Algorithms to Neural Networks

552

(2)

Fuzzy Genetic Modeling

554

(2)

Use of Genetic Algorithms in the Design of Neural Networks

556

(5)

References

557

(2)

Problems

559

(2)

Epilogue

561

(6)

Introduction

561

(1)

Is Artificial Intelligence Really Intelligent?

562

(1)

The Role of Neurofuzzy Technology

563

(1)

Last Thoughts

564

(3)

References

565

(2)

Appendix: T Norms and S Norms

567

(8)

Index

575

What is included with this book?

The New copy of this book will include any supplemental materials advertised. Please check the title of the book to determine if it should include any access cards, study guides, lab manuals, CDs, etc.

The Used, Rental and eBook copies of this book are not guaranteed to include any supplemental materials. Typically, only the book itself is included. This is true even if the title states it includes any access cards, study guides, lab manuals, CDs, etc.

FREE SHIPPING

Amazon no longer offers textbook rentals. We do!

Fuzzy and Neural Approaches in Engineering

Summary

Author Biography

Table of Contents

Supplemental Materials