Engineering Applications of Soft Computing

This book bridges the gap between Soft Computing techniques and their applications to complex engineering problems. In each chapter we endeavor to explain the basic ideas behind the proposed applications in an accessible format for readers who may not possess a background in some of the fields. Therefore, engineers or practitioners who are not familiar with Soft Computing methods will appreciate that the techniques discussed go beyond simple theoretical tools, since they have been adapted to solve significant problems that commonly arise in such areas. At the same time, the book will show members of the Soft Computing community how engineering problems are now being solved and handled with the help of intelligent approaches. Highlighting new applications and implementations of Soft Computing approaches in various engineering contexts, the book is divided into 12 chapters. Further, it has been structured so that each chapter can be read independently of the others. Preface 6 Contents 11 1 Introduction 16 1.1 Soft Computing 16 1.2 Fuzzy Logic 17 1.3 Neural Networks 18 1.4 Evolutionary Computation 18 1.5 Definition of an Optimization Problem 19 1.6 Classical Optimization 21 1.7 Optimization with Evolutionary Computation 23 1.8 Soft Computing in Engineering 25 References 26 2 Motion Estimation Algorithm Using Block-Matching and Harmony Search Optimization 28 2.1 Introduction 29 2.2 Harmony Search Algorithm 32 2.2.1 The Harmony Search Algorithm 32 2.2.1.1 Initializing the Problem and Algorithm Parameters 32 2.2.1.2 Harmony Memory Initialization 33 2.2.1.3 Improvisation of New Harmony Vectors 33 2.2.1.4 Updating the Harmony Memory 34 2.2.2 Computational Procedure 34 2.3 Fitness Approximation Method 35 2.3.1 Updating the Individual Database 35 2.3.2 Fitness Calculation Strategy 35 2.3.3 HS Optimization Method 38 2.4 Motion Estimation and Block-Matching 39 2.5 Block-Matching Algorithm Based on Harmony Search with the Estimation Strategy 41 2.5.1 Initial Population 42 2.5.2 Tuning of the HS Algorithm 44 2.5.3 The HS-BM Algorithm 44 2.5.4 Discussion on the Accuracy of the Fitness Approximation Strategy 46 2.6 Experimental Results 49 2.6.1 HS-BM Results 49 2.6.2 Results on H.264 51 2.7 Conclusions 56 References 57 3 Gravitational Search Algorithm Applied to Parameter Identification for Induction Motors 60 3.1 Introduction 60 3.2 Problem Statement 62 3.3 Gravitational Search Algorithm 64 3.4 Experimental Results 65 3.4.1 Induction Motor Parameter Identification 66 3.4.2 Statistical Analysis 70 3.5 Conclusions 70 References 71 4 Color Segmentation Using LVQ Neural Networks 73 4.1 Introduction 73 4.1.1 Histogram Thresholding and Color Space Clustering 74 4.1.2 Edge Detection 75 4.1.3 Probabilistic Methods 75 4.1.4 Soft-Computing Techniques 76 4.1.5 Scheme 76 4.2 Background Issues 77 4.2.1 RGB Space Color 77 4.2.2 Artificial Neural Networks 77 4.3 Competitive Networks 78 4.4 Learning Vectors Quantization Vectors 80 4.5 Architecture of the Color Segmentation System 80 4.6 Implementation 82 4.7 Results and Discussion 84 4.8 Conclusions 88 References 88 5 Global Optimization Using Opposition-Based Electromagnetism-Like Algorithm 90 5.1 Introduction 90 5.2 Electromagnetism: Like Optimization Algorithm (EMO) 93 5.2.1 Initialization 93 5.2.2 Local Search 94 5.2.3 Total Force Vector Computation 95 5.2.4 Movement 96 5.3 Opposition-Based Learning (OBL) 97 5.3.1 Opposite Number 97 5.3.2 Opposite Point 97 5.3.3 Opposite-Based Optimization 97 5.4 Opposition-Based Electromagnetism-Like Optimization Algorithm 98 5.4.1 Opposition-Based Population Initialization 98 5.4.2 Opposition-Based Production for New Generation 100 5.5 Experimental Results 101 5.5.1 Test Problems 102 5.5.2 Parameter Settings for the Involved EMO Algorithms 102 5.5.3 Results 106 5.6 Conclusions 111 References 111 6 Multi-threshold Segmentation Using Learning Automata 114 6.1 Introduction 114 6.2 Gaussian Approximation 117 6.3 Learning Automata (LA) 118 6.3.1 CARLA Algorithm 120 6.4 Implementation 122 6.5 Experimental Results 124 6.5.1 LA Algorithm Performance in Image Segmentation 124 6.5.2 Comparing the LA Algorithm Versus the EM and LM Methods 128 6.6 Conclusions 132 References 139 7 Real-Time Gaze Control Using Neurofuzzy Prediction System 141 7.1 Introduction 141 7.2 Adaptive Neurofuzzy Inference System 142 7.2.1 ANFIS Architecture 143 7.2.1.1 Layer 1 144 7.2.1.2 Layer 2 145 7.2.1.3 Layer 3 145 7.2.1.4 Layer 4 145 7.2.1.5 Layer 5 145 7.2.2 Hybrid Learning Algorithm 146 7.3 Implementation 146 7.3.1 Description 146 7.3.2 Segmentation Algorithm and Localization 147 7.3.3 Controller 148 7.3.4 Predictor Design 150 7.4 Results 152 7.5 Conclusions 153 References 153 8 Clonal Selection Algorithm Applied to Circle Detection 154 8.1 Introduction 154 8.2 Clonal Selection Algorithm 157 8.2.1 Definitions 158 8.2.2 CSA Operators 158 8.2.2.1 Clonal Proliferation Operator ( T_{{\rm P}}^{{\rm C}} ) 159 8.2.2.2 Affinity Maturation Operator ( T_{M}^{{\rm A}} ) 159 8.2.2.3 Clonal Selection Operator ( T_{{\rm S}}^{{\rm C}} ) 160 8.3 Circle Detection Using CSA 162 8.3.1 Individual Representation 162 8.3.1.1 Objective Function or Matching Function 164 8.3.1.2 Implementation of CSA 164 8.4 Experimental Results 166 8.4.1 Parametric Setup 166 8.4.2 Error Score and Success Rate 166 8.4.2.1 Presentation of Results 167 8.5 Conclusions 172 References 173 9 States of Matter Algorithm Applied to Pattern Detection 176 9.1 Introduction 176 9.2 Gaussian Approximation 179 9.3 States of Matter Search (SMS) 180 9.3.1 Definition of Operators 180 9.3.1.1 Direction Vector 180 9.3.1.2 Collision 182 9.3.1.3 Random Positions 182 9.3.1.4 Best Element Updating 183 9.3.2 SMS Algorithm 183 9.3.2.1 General Procedure 183 9.3.2.2 The Complete Algorithm 183 9.4 Fitness Approximation Method 187 9.4.1 Updating Individual Database 187 9.4.2 Fitness Calculation Strategy 188 9.4.3 Proposed Optimization SMS Method 190 9.5 Pattern Detection Process 191 9.6 PD Algorithm Based on SMS with the Estimation Strategy 193 9.6.1 The SMS-PD Algorithm 194 9.7 Experimental Results 196 9.8 Conclusions 201 References 202 10 Artificial Bee Colony Algorithm Applied to Multi-threshold Segmentation 204 10.1 Introduction 204 10.2 Gaussian Approximation 206 10.3 Artificial Bee Colony (ABC) Algorithm 207 10.3.1 Biological Bee Profile 208 10.3.2 Description of the ABC Algorithm 208 10.3.2.1 Initializing the Population 208 10.3.2.2 Send Employed Bees 209 10.3.2.3 Select the Food Sources by the Onlooker Bees 209 10.3.2.4 Determine the Scout Bees 210 10.4 Determination of Thresholding Values 210 10.5 Experimental Results 211 10.5.1 Comparing the ABC Algorithm Versus the EM and LM Methods 213 10.6 Conclusions 224 References 224 11 Learning Automata Applied to Planning Control 226 11.1 Introduction 226 11.2 Planning Strategy Design 229 11.2.1 Closed-Loop Planning Configuration 230 11.2.2 Models and Projections into the Future 231 11.2.3 Optimization and Method for Plan Selection 232 11.3 Learning Automata 234 11.3.1 CARLA Algorithm 235 11.4 Implementation 237 11.4.1 Level Control in a Surge Tank 238 11.4.2 Planning System 239 11.4.3 LA Optimization 240 11.5 Results 242 11.6 Conclusions 246 References 247 12 Fuzzy-Based System for Corner Detection 249 12.1 Introduction 249 12.2 Fuzzy Rule-Based System 251 12.2.1 Fuzzy System 251 12.2.2 Robustness 256 12.2.3 Corner Selection 258 12.3 Experimental Results 258 12.4 Performance Comparison 260 12.4.1 Stability Criterion 262 12.4.2 Noise Immunity 263 12.4.3 Computational Effort 264 12.4.4 Comparison Results 264 12.5 Conclusions 265 References 267 Front Matter....Pages i-xv Introduction....Pages 1-12 Motion Estimation Algorithm Using Block-Matching and Harmony Search Optimization....Pages 13-44 Gravitational Search Algorithm Applied to Parameter Identification for Induction Motors....Pages 45-57 Color Segmentation Using LVQ Neural Networks....Pages 59-75 Global Optimization Using Opposition-Based Electromagnetism-Like Algorithm....Pages 77-100 Multi-threshold Segmentation Using Learning Automata....Pages 101-127 Real-Time Gaze Control Using Neurofuzzy Prediction System....Pages 129-141 Clonal Selection Algorithm Applied to Circle Detection....Pages 143-164 States of Matter Algorithm Applied to Pattern Detection....Pages 165-192 Artificial Bee Colony Algorithm Applied to Multi-threshold Segmentation....Pages 193-214 Learning Automata Applied to Planning Control....Pages 215-237 Fuzzy-Based System for Corner Detection....Pages 239-258