- ホーム
- > 洋書
- > 英文書
- > Science / Mathematics
Full Description
Thorough review of how artificial intelligence can enhance the design, control, and optimization of power electronics systems
Artificial Intelligence for Power Electronics provides a comprehensive overview of the intersection between artificial intelligence (AI) and the field of power electronics, exploring how AI can revolutionize and enhance the design, control, and optimization of power electronics systems. The book covers the fundamentals of AI, the fundamentals of power electronics and the challenges the field faces in design to production, and the solutions of these challenges through AI methods. Example solutions, along with Q&A review sections, are included throughout the text, with coverage of both Python and MATLAB.
Topics discussed in Artificial Intelligence for Power Electronics include:
Supervised, unsupervised, and reinforcement machine learning and the role of data in training machine learning models
Techniques for AI data collection in power electronics and how to clean, normalize, and handle missing values of data
Optimization techniques such as Particle Swarm Optimization and Ant Colony Optimization
Detection techniques for identifying faults and anomalies and clustering algorithms to group similar operational behavior
Essential Python libraries for machine learning and how to perform machine learning on a Raspberry Pi
Delivering an industry-specific approach to AI applications, Artificial Intelligence for Power Electronics is a helpful reference for undergraduate, postgraduate, and PhD students in electrical, electronic, and computer engineering. Mechanical engineers and other industry professionals may also find it valuable.
Contents
About the Editors xvii
List of Contributors xix
Preface xxi
1 Fundamentals of Power Electronics and Key Challenges 1
Azra Malik and Ahteshamul Haque
1.1 Introduction 1
1.2 Fundamental Concepts and Definitions 4
1.2.1 Uncontrolled Switches 5
1.2.2 Semi-Controlled Switches 5
1.2.3 Fully Controlled Switch 7
1.2.3.1 Bipolar Junction Transistor (BJT) 7
1.2.3.2 Metal Oxide Semiconductor Field Effect Transistor (MOSFET) 9
1.2.3.3 The Insulated Gate Bipolar Transistor (IGBT) 10
1.2.3.4 Gate Turn-Off (GTO) Thyristor 11
1.2.3.5 MOS-Controlled Thyristor (MCT) 11
1.3 Fundamental Principles Related with Power Electronic Converters 13
1.3.1 AC/DC Converters or Rectifiers 13
1.3.2 DC/DC Converters 15
1.3.3 DC/AC Converters 19
1.3.3.1 Single-Phase Inverter 19
1.3.3.2 Three-Phase Inverters 21
1.3.4 AC/AC Converters 22
1.4 Case Study 22
1.4.1 AC/DC Converter (Rectifier) 23
1.4.2 DC/DC Converter (Buck Converter) 23
1.4.3 DC/AC Inverter 23
1.5 Challenges in Power Electronics 24
1.5.1 Renewable Energy Integration Challenges 24
1.5.2 Challenges Arising from Power Electronics Integration 25
1.6 Future Trends in Power Electronics 26
1.6.1 Wide-Bandgap (WBG) Semiconductors 26
1.6.2 Trends in Materials and Packaging for Better Performance 27
1.6.3 Role of Power Electronics in Sustainable Energy Systems 27
1.7 Conclusion 28
Exercises 29
References 30
2 Introduction of AI and Utility for Power Electronics Applications 33
Suwaiba Mateen and Ahteshamul Haque
2.1 Introduction 33
2.2 Intersection of Artificial Intelligence and Power Electronics 35
2.3 AI Techniques in Power Electronics 37
2.3.1 Expert Systems (ES) 37
2.3.2 Fuzzy Logic (FL) 38
2.3.3 Metaheuristic Methods 39
2.3.3.1 Trajectory-Based Methods 39
2.3.3.2 Population-Based Methods 39
2.3.4 Machine Learning 40
2.3.4.1 Supervised Learning 40
2.3.4.2 Unsupervised Learning 42
2.3.4.3 Reinforcement Learning 42
2.4 Applications of AI in Power Electronics 46
2.4.1 Design Using AI 46
2.4.2 Control Using AI 47
2.4.3 Maintenance Using AI 48
2.5 Case Studies and Real-World Examples 49
2.5.1 Modeling of a Non-linear System Using NN 49
2.5.2 Control of a Boost Converter Using ANFIS 51
2.5.3 Classification of Faults in Grid Tied PV System Using kNN 55
2.6 Challenges and Limitations 57
2.7 Conclusion 59
Exercises 60
References 61
3 Machine Learning Fundamentals 67
Ahteshamul Haque, Azra Malik, and Mansha Khursheed
3.1 Introduction 67
3.2 Key Components of Machine Learning 70
3.2.1 Data Preparation 70
3.2.2 Feature Selection and Engineering 71
3.2.3 Model Selection 72
3.2.4 Model Training 72
3.2.5 Model Evaluation 75
3.3 Fundamental Concepts and Definitions 76
3.3.1 Regression 77
3.3.2 Classification 79
3.3.3 Clustering 81
3.3.4 Embedding Algorithms 81
3.4 Machine Learning (ML) Applications in Power Electronics 82
3.4.1 ml Application in Forecasting 82
3.4.2 ml Application in Control 84
3.4.3 ml Application in Electric Vehicle (EV) 86
3.4.4 ml Application in Maintenance 87
3.5 Case Study 89
3.5.1 Regression (PV Generation Forecasting) 89
3.5.2 Classification (Inverter Fault Classification) 91
3.6 Challenges 95
3.7 Future Research Directions 96
3.8 Conclusion 97
Exercises 97
References 98
4 Data Collection and Pre-processing 105
Manauwar Hussain, Suwaiba Mateen, and Ahteshamul Haque
4.1 Introduction 105
4.2 Data Collection in Power Electronics 106
4.2.1 Types of Data in Power Electronics 107
4.2.2 Data Sources 108
4.2.3 Data Collection Methods 109
4.3 Data Quality and Challenges 110
4.4 Data Pre-processing Techniques 111
4.4.1 Data Cleaning 112
4.4.2 Data Transformation 113
4.4.3 Feature Engineering 115
4.4.4 Data Reduction Techniques 115
4.5 Data Annotation and Labeling 117
4.5.1 Methods for Data Labeling 118
4.6 Case Study: Data Smoothing and Detecting Outliers 119
4.6.1 Moving Window Methods 120
4.6.2 Common Smoothing Methods 122
4.6.3 Detecting Outliers 125
4.6.4 Nonuniform Data 125
4.7 Challenges and Limitations 129
4.8 Conclusion 129
Exercises 130
References 131
5 Fuzzy Logic and Metaheuristic Methods in Power Electronics 137
Fatima Shabir Zehgeer and Ahteshamul Haque
5.1 Introduction 137
5.2 Applications of Fuzzy Logic Methods in Power Electronics 139
5.2.1 Converter Control 139
5.2.2 Power Quality Improvement 141
5.2.3 Fault Detection in Power Electronic Converters 142
5.2.4 Power Flow Control 143
5.2.5 Design of Power Electronic Devices 143
5.3 Applications of Metaheuristic Methods in Power Electronics 143
5.3.1 Power Quality Improvement 144
5.3.2 Control 144
5.3.3 Optimal Design of the Power Electronic Converters 144
5.4 Hybrid Approaches: Fuzzy Logic and Metaheuristic Methods in Power Electronics 145
5.4.1 Control of the Power Electronic Converters 145
5.4.2 Power Quality Improvement 147
5.4.3 Power Flow Control 147
5.4.4 Design 149
5.5 Case Studies and Real-World Examples 149
5.5.1 Fuzzy Logic Control of the Buck Converter 149
5.5.2 Particle Swarm Optimization-Based MPPT Control Under Different Irradiances 151
5.5.3 Fuzzy-Logic-Control-Based MPPT of the PV System 157
5.6 Conclusion 162
Exercises 165
References 166
6 Supervised Learning for Power Electronics 173
Md Zafar Khan and Ahteshamul Haque
6.1 Introduction 173
6.2 Types of Supervised Learning 174
6.2.1 Support Vector Machines (SVMs) 174
6.2.2 Neural Networks 177
6.2.3 Decision Trees 178
6.3 Applications in Power Electronics 182
6.3.1 Regression Applications 182
6.3.1.1 Energy Consumption Forecasting 182
6.3.1.2 Power Output Prediction 183
6.3.1.3 Converter Control 184
6.3.2 Classification Applications 184
6.3.2.1 Fault Location and Type Detection 185
6.3.2.2 Islanding Detection 188
6.3.3 Clustering Applications 188
6.3.3.1 Load Profiling 188
6.3.3.2 Anomaly Detection 190
6.4 Case Study: Predicting Power Consumption in an Electric Motor Using Support Vector Regression (SVR) in MATLAB 190
6.5 Challenges and Future Prospects 196
6.6 Conclusion 196
Exercises 196
References 198
7 Unsupervised Learning for Anomaly Detection 201
Ahteshamul Haque and Mohammed Ali Khan
7.1 Introduction 201
7.2 Faults in Power Electronics 202
7.2.1 Open Circuit Faults 203
7.2.2 Short Circuit Faults 204
7.2.3 Gate Drive Faults 204
7.2.4 Thermal Faults 205
7.2.5 Component Aging and Degradation 205
7.2.6 Electromagnetic Interference (EMI) 206
7.3 Unsupervised Learning 206
7.3.1 Statistical Methods 207
7.3.1.1 Z-Score 208
7.3.1.2 Isolation Forest 208
7.3.2 Clustering-Based Methods 209
7.3.2.1 k-Means 209
7.3.2.2 Dbscan 210
7.3.3 Dimensionality Reduction-Based Methods 210
7.3.3.1 Principal Component Analysis (PCA) 211
7.3.3.2 Autoencoders 211
7.3.4 Proximity-Based Methods 212
7.3.4.1 k-Nearest Neighbors (k-NNs) 212
7.3.4.2 Local Outlier Factor (LOF) 213
7.4 Modeling System for the Case Study 214
7.5 Conclusion 221
Exercises 221
References 222
8 Reinforcement Learning and Control 229
Azra Malik, Suwaiba Mateen, and Ahteshamul Haque
8.1 Introduction 229
8.2 Basics of Reinforcement Learning (RL) 231
8.2.1 Markov Decision Process (MDP) 231
8.2.2 Bellman Equation 232
8.2.3 On-policy and Off-policy Methods 233
8.2.4 Dynamic Programming Methods 233
8.2.5 Monte Carlo Method 234
8.2.6 Temporal Difference Method 234
8.2.7 Policy Gradient Method 234
8.3 RL Methods 235
8.3.1 Value-based DRL 236
8.3.1.1 Deep Q-learning 236
8.3.1.2 Double Deep Q-learning 237
8.3.1.3 Dueling Deep Q-learning 238
8.3.1.4 Fitted Q-learning 238
8.3.2 Policy-based DRL 239
8.3.2.1 Trust-region Optimization (TRO) 239
8.3.2.2 Soft Actor-Critic (SAC) 240
8.3.2.3 Advantage Actor-Critic (AAC) 241
8.3.2.4 Deep Deterministic Policy Gradient (DDPG) 241
8.3.3 Multi-agent DRL 242
8.4 Reinforcement Learning in Power Electronics Applications 243
8.4.1 Control Optimization of Power Converters 243
8.4.2 Energy Management 247
8.4.3 Demand-Side Management 249
8.4.4 Grid Stability 251
8.5 Case Study - RL-based Control of Buck Converter 251
8.5.1 Problem Statement and Significance 251
8.5.2 Methodology 252
8.5.2.1 Reinforcement Learning Framework 252
8.6 Future Research Directions 259
8.7 Conclusion 259
Exercises 260
References 261
9 Implementation of Machine Learning for Power Electronics Application Using MATLAB 267
Manauwar Hussain, Ahteshamul Haque, and Md Zafar Khan
9.1 Introduction 267
9.2 Machine Learning 269
9.2.1 Being Familiar with Machine Learning 269
9.2.2 Importance of Machine Learning 270
9.3 Types of Machine Learning 272
9.3.1 Supervised Learning 272
9.3.2 Unsupervised Learning 273
9.3.3 Semi-supervised Learning 273
9.3.4 Reinforcement Learning 274
9.4 ml in Power Electronics 275
9.5 Current Trends and Research in the Integration of ML with Power Electronics 276
9.5.1 Control Techniques in Power Electronics 276
9.5.1.1 Linear Control Techniques 276
9.5.1.2 Non-linear Control Techniques 278
9.6 Machine Learning in Power Electronics Using MATLAB 280
9.6.1 Overview of MATLAB 280
9.6.2 MATLAB Toolboxes for ML and Power Electronics 280
9.6.2.1 Data Preparation and Pre-processing 284
9.6.2.2 Feature Selection and Extraction 284
9.6.2.3 Developing ML Models in MATLAB 285
9.7 Case Study 285
9.7.1 Case Study 1: Predicting Efficiency of a DC-DC Converter 285
9.7.2 Case Study 2: Predicting the Temperature of a Power Semiconductor Device Using a Neural Network 291
9.8 Conclusion 297
Exercises 297
References 299
10 Implementation of Machine Learning for Power Electronics Application Using PYTHON 301
Mohammad Amir, Izhar Ahmad Saifi, and Ahteshamul Haque
10.1 Introduction 301
10.1.1 Supervised Learning 301
10.1.2 Unsupervised Learning 302
10.1.3 Reinforcement Learning (RL) 303
10.1.4 Deep Learning for Fault Diagnosis 304
10.1.5 Hybrid Approach 306
10.2 ml Algorithms Used in Power Electronics Utilizing PYTHON Platform 306
10.2.1 Key Domains of ML in Power Electronics Using PYTHON 307
10.3 PYTHON Library and Model Development 308
10.3.1 Extensive ML Libraries and PYTHON Framework 309
10.3.2 Cross-Platform and Hardware Compatibility 310
10.4 Stepwise Developing a Power Electronics Classification Model in Python 310
10.4.1 PYTHON Used to Control 30-ph Grid Connected System 313
10.5 Development of ML Classification Model Using PYTHON for PEs Converters 315
10.5.1 Case Study: Fault Classification Using ML (SVM and KNN) Techniques 316
10.6 Challenges of Utilizing ML with Python for PEs Applications 321
10.7 Conclusion and Future Scope 322
Exercises 323
References 324
11 Integration of AI in Power Electronics in Real-time 329
Kurukuru Varaha Satya Bharath and Ahteshamul Haque
11.1 Overview 329
11.2 Control Development 330
11.2.1 Electro-Thermal Modeling 330
11.2.2 Thermal Model 333
11.2.3 Thermal Observer 337
11.2.4 Methods for Active Loss Manipulation 341
11.3 Overview of Rapid Control Prototyping (RCP) 344
11.3.1 Imperix B-Box 345
11.3.2 ACG-SDK in MATLAB 347
11.4 System Configuration 348
11.4.1 Description of the 3-Phase Dual Converter System 348
11.4.2 Integration with FOC and DPWM 348
11.5 Development Process 350
11.5.1 System Design in HIL 350
11.5.1.1 Integration of Efficiency Models into System Simulation 353
11.5.2 Implementation in Imperix B-Box 353
11.5.2.1 Control Strategy Implementation 355
11.6 Hardware-in-the-Loop (HIL) and RCP Interface 358
11.6.1 HIL-RCP Integration 358
11.6.2 Results and Observations 359
11.7 Conclusion 361
Exercises 364
References 365
Index 369
