Optimization, Learning, and Control for Interdependent Complex Networks (eBook)

M. Hadi Amini (Herausgeber)

eBook Download: PDF

2020 | 1st ed. 2020
X, 304 Seiten
Springer International Publishing (Verlag)
978-3-030-34094-0 (ISBN)

Infos

This book focuses on a wide range of optimization, learning, and control algorithms for interdependent complex networks and their role in smart cities operation, smart energy systems, and intelligent transportation networks. It paves the way for researchers working on optimization, learning, and control spread over the ?elds of computer science, operation research, electrical engineering, civil engineering, and system engineering. This book also covers optimization algorithms for large-scale problems from theoretical foundations to real-world applications, learning-based methods to enable intelligence in smart cities, and control techniques to deal with the optimal and robust operation of complex systems. It further introduces novel algorithms for data analytics in large-scale interdependent complex networks.

 •  Speci?es the importance of efficient theoretical optimization and learning methods in    dealing with emerging problems in the context of interdependent networks

 •  Provides a comprehensive investigation of advance data analytics and machine learning algorithms for large-scale complex networks

 •  Presents basics and mathematical foundations needed to enable efficient decision making and intelligence in interdependent complex networks

 

M. Hadi Amini is an Assistant Professor at the School of Computing and Information Sciences at Florida International University (FIU). He is also the founding director of Sustainability, Optimization, and Learning for InterDependent networks laboratory (solid lab). He received his Ph.D. and M.Sc. from Carnegie Mellon University in 2019 and 2015 respectively. He also holds a doctoral degree in Computer Science and Technology. Prior to that, he received M.Sc. from Tarbiat Modares University in 2013, and the B.Sc. from Sharif University of Technology in 2011.

M. Hadi Amini is an Assistant Professor at the School of Computing and Information Sciences at Florida International University (FIU). He is also the founding director of Sustainability, Optimization, and Learning for InterDependent networks laboratory (solid lab). He received his Ph.D. and M.Sc. from Carnegie Mellon University in 2019 and 2015 respectively. He also holds a doctoral degree in Computer Science and Technology. Prior to that, he received M.Sc. from Tarbiat Modares University in 2013, and the B.Sc. from Sharif University of Technology in 2011. His research interests include distributed machine learning and optimization algorithms, distributed intelligence, sensor networks, interdependent networks, and cyberphysical resilience. Application domains include energy systems, healthcare, device-free human sensing, and transportation networks. 

Prof. Amini is a life member of IEEE-Eta Kappa Nu (IEEE-HKN), the honor society of IEEE. He organized a panel on distributed learning and novel artificial intelligence algorithms, and their application to healthcare, robotics, energy cybersecurity, distributed sensing, and policy issues in 2019 workshop on artificial intelligence at FIU. He also served as President of Carnegie Mellon University Energy Science and Innovation Club; as technical program committee of several IEEE and ACM conferences; and as the lead editor for a book series on ''Sustainable Interdependent Networks'' since 2017. He has published more than 80 refereed journal and conference papers, and book chapters. He has co-authored two books, and edited three books on various aspects of optimization and machine learning for interdependent networks. He is the recipient of the best paper award of 'IEEE Conference on Computational Science & Computational Intelligence' in 2019, best reviewer award from four IEEE Transactions, the best journal paper award in 'Journal of Modern Power Systems and Clean Energy', and the dean's honorary award from the President of Sharif University of Technology.

Inhalt

Preface 6
Contents 7
About the Editor 9
1 Panorama of Optimization, Control, and Learning Algorithms for Interdependent SWEET (Societal, Water, Energy, Economic, and Transportation) Networks 11
1.1 Introduction 11
1.2 Part I: Theoretical Algorithms for Optimization, Learning, and Data Analytics in Interdependent Complex Networks 14
1.2.1 Chapter 2: Promises of Fully Distributed Optimization for IoT-Based Smart City Infrastructures: Theory and Applications 14
1.2.2 Chapter 3: Evolutionary Computation, Optimization, and Learning Algorithms for Data Science 14
1.2.3 Chapter 4: Applications of Nature-Inspired Algorithms for Dimension Reduction: Enabling Efficient Data Analytics 15
1.2.4 Chapter 5: Feature Selection in High-Dimensional Data 16
1.2.5 Chapter 6: An Introduction to Advanced Machine Learning: Meta-Learning Algorithms, Applications, and Promises 17
1.3 Part II: Application of Optimization, Learning, and Control in Interdependent Complex Networks 17
1.3.1 Chapter 7: Predictive Analytics in Future Power Systems: A Panorama and State-of-the-Art of Deep Learning Applications 17
1.3.2 Chapter 8: Bilevel Adversary-Operator Cyberattack Framework and Algorithms for Transmission Networks in Smart Grids 18
1.3.3 Chapter 9: Toward Operational Resilience of Smart Energy Networks in Complex Infrastructures 18
1.3.4 Chapter 10: Control of Cooperative Unmanned Aerial Vehicles: Review of Applications, Challenges, and Algorithms 19
1.3.5 Chapter 11: An Optimal Approach for Load-Frequency Control of Islanded Microgrids Based on Non-linear Model 19
1.3.6 Chapter 12: PV Design for Smart Cities and Demand Forecasting Using Truncated Conjugate Gradient Algorithm 20
References 21
Part I Theoretical Algorithms for Optimization, Learning, and Data Analytics in Interdependent Complex Networks 22
2 Promises of Fully Distributed Optimization for IoT-Based Smart City Infrastructures 23
2.1 Introduction 24
2.1.1 Motivation 24
2.1.2 Related Works 25
2.1.3 Contribution 27
2.1.4 Organization 28
2.2 A Novel Holistic Framework for Interdependent Operation of Power Systems and Electrified Transportation networks 28
2.3 Definition of Agents and Their Corresponding Features 30
2.3.1 Power System-Specific Agents 31
2.3.2 Transportation Network-Specific Agents 32
2.3.3 Coupling Agents 32
2.4 General Optimization Problem 35
2.4.1 Problem Formulation 35
2.4.2 Optimality Conditions 35
2.5 Consensus+Innovations Based Distributed Algorithm 36
2.5.1 Distributed Decision Making: General Distributed Update Rule 36
2.5.2 Agent-Based Distributed Algorithm 36
2.6 Conclusions 37
Appendix 1: Convergence Analysis 38
References 39
3 Evolutionary Computation, Optimization, and Learning Algorithms for Data Science 44
3.1 Introduction 45
3.1.1 Overview 45
3.1.2 Motivation 46
3.1.3 Curse of Dimensionality 48
3.1.4 Nature-Inspired Computation 48
3.1.5 Nature-Inspired Meta-Heuristic Computation 49
3.1.6 Nature-Inspired Evolutionary Computation 49
3.1.6.1 Evolutionary-Based Memetic Algorithms 49
3.1.7 Organization 50
3.2 Feature Extraction Techniques 51
3.3 Bio-Inspired Evolutionary Computation 53
3.3.1 Overview of Evolutionary Algorithms 53
3.3.2 Genetic Algorithm vs. Genetic Programming 56
3.3.2.1 Genetic Algorithm 56
3.3.2.2 Genetic Programming 58
3.3.3 Artificial Bee Colony Algorithm 60
3.3.4 Particle Swarm Optimization Algorithm 61
3.3.5 Ant Colony Optimization (ACO) 63
3.3.6 Grey Wolf Optimizer (GWO) 64
3.3.7 Coyote Optimization Algorithm (COA) 65
3.3.8 Other Optimization Algorithms 65
3.4 Conclusion 67
References 68
4 Applications of Nature-Inspired Algorithms for Dimension Reduction: Enabling Efficient Data Analytics 73
4.1 Introduction 74
4.1.1 Overview 74
4.1.2 Organization 75
4.2 Application of Evolutionary Algorithms 75
4.2.1 Feature Extraction Optimization 78
4.2.1.1 Feature Selection for Image Classification 78
4.2.2 Feature Selection for Network Traffic Classification 83
4.2.3 Feature Selection Benchmarks 85
4.3 Discussion 86
4.4 Conclusion 86
References 87
5 Feature Selection in High-Dimensional Data 91
5.1 Overview 92
5.2 Intrinsic Characteristics of High-Dimensional Data 93
5.2.1 Large Number of Features 93
5.2.2 Small Number of Samples 94
5.2.3 Class Imbalance 94
5.2.4 Label Noise 96
5.2.5 Intrinsic Characteristics of Microarray Data 97
5.3 Feature Selection 98
5.4 Filter Methods 99
5.4.1 Similarity-Based Methods 99
5.4.1.1 Relief and ReliefF 100
5.4.1.2 Fisher Score 100
5.4.1.3 Laplacian Score 101
5.4.2 Statistical-Based Methods 101
5.4.2.1 Correlation-Based Feature Selection (CFS) 102
5.4.2.2 Low Variance 102
5.4.2.3 T-Score 102
5.4.2.4 Information Theoretical-Based Methods 103
5.4.2.5 FCBF 103
5.4.2.6 Minimum-Redundancy-Maximum-Relevance (mRMR) 103
5.4.2.7 Information Gain 104
5.5 Wrapper Methods 104
5.5.1 ABACOH and ACO 105
5.5.2 PSO 107
5.5.3 IBGSA 108
5.6 Hybrid Method 110
5.7 Embedded Methods 111
5.8 Ensemble Techniques 112
5.9 Practical Evaluation 117
5.9.1 Dataset 117
5.9.2 Performance Evaluation Criteria 117
5.9.3 Data Normalization 119
5.9.4 Analysis of Filter Algorithms 119
5.9.5 Analysis of Hybrid-Ensemble Methods 122
5.9.5.1 Hybrid-Ensemble 1 122
5.9.5.2 Hybrid-Ensemble 2 126
5.10 Summary 128
References 130
6 An Introduction to Advanced Machine Learning: Meta-Learning Algorithms, Applications, and Promises 135
6.1 Introduction 136
6.2 Machine Learning: Challenges and Drawbacks 137
6.3 Meta-Learning Algorithms 139
6.3.1 Model-Based MTL 139
6.3.2 Metric-Based Learning 140
6.3.3 Gradient Decent-Based Learning 141
6.4 Promises of Meta-Learning 141
6.4.1 Few-Shot Learning 144
6.4.2 One-Shot Learning 145
6.4.3 Zero-Shot Learning 145
6.5 Discussion 146
6.6 Conclusion 147
References 148
Part II Application of Optimization, Learning and Control in Interdependent Complex Networks 151
7 Predictive Analytics in Future Power Systems: A Panorama and State-Of-The-Art of Deep Learning Applications 152
7.1 Introduction 153
7.1.1 Motivation 154
7.1.2 Classification of Power Systems Forecasting Models 155
7.1.2.1 Classification Based on the Domain of Application in Power Systems 156
7.1.2.2 Classification Based on Timescale 159
7.1.3 Organization of the Chapter 160
7.2 Forecasting in Power Systems Using Classical Approaches 161
7.2.1 Time Series Data 161
7.2.2 Statistical Forecasting Approaches 163
7.2.2.1 Naïve Model Approach 163
7.2.2.2 Exponential Smoothing 164
7.2.2.3 Autoregressive Moving Average (ARMA) Models 164
7.2.2.4 Autoregressive Moving Integrated Average (ARIMA) Models 166
7.2.3 Machine Learning Forecasting Approaches 166
7.2.3.1 Support Vector Regression 167
7.2.3.2 Gaussian Process Regression 168
7.2.4 Shortcomings of Classical Approaches 169
7.3 Forecasting in Power Systems Using Deep Learning 169
7.3.1 Deep Learning 169
7.3.1.1 Recurrent Neural Network 170
7.3.1.2 Long Short-Term Memory Network 171
7.3.1.3 Other Relevant Models 173
7.3.2 Deep Learning Applications 173
7.3.2.1 Load Forecasting 173
7.3.2.2 Generation Forecasting 174
7.3.2.3 Electricity Price Forecasting and Electric Vehicle Charging 174
7.3.3 Deep Learning Strengths and Shortcomings 174
7.3.3.1 Strengths 175
7.3.3.2 Shortcomings 175
7.4 Case Study: Multi-Timescale Solar Irradiance Forecasting Using Deep Learning 175
7.4.1 Data 176
7.4.1.1 Global Horizontal Irradiance 177
7.4.1.2 Exogenous Input Variables 177
7.4.1.3 Data Preprocessing and Postprocessing 178
7.4.2 Model Architecture and Training 178
7.4.3 Results 179
7.4.3.1 Single Time Horizon Model 179
7.4.3.2 Multi-Time-Horizon Model 179
7.5 Summary and Future Work 182
7.5.1 Deterministic Versus Probabilistic Forecasting 182
7.5.2 Other Potential Applications 183
References 183
8 Bi-level Adversary-Operator Cyberattack Framework and Algorithms for Transmission Networks in Smart Grids 188
8.1 Introduction 189
8.1.1 Overview 189
8.2 DC Power Flow Model 191
8.3 False Data Injection Attacks Based on DC State Estimation 193
8.4 Attacker's Problem: Finding the Optimal Set of Target Transmission Lines using MILP 195
8.4.1 Identifying Feasible Attacks 198
8.5 Operator's Problem: Bad Data Detection to Prevent Outages Caused by Cyberattack 198
8.6 Case Studies 201
8.6.1 Feasibility of Line Overflow 202
8.6.2 Targeted Attack on Line 15 203
8.6.3 Severe Attack on an Area 205
8.7 Conclusion 205
References 206
9 Toward Operational Resilience of Smart Energy Networks in Complex Infrastructures 208
9.1 Introduction 209
9.1.1 Overview 209
9.2 Resilience Enhancement Scheme 210
9.3 Real-Time Decision Making Process 212
9.4 Optimization Model 213
9.4.1 Pre-event Preparation Strategy 213
9.4.2 Mid-Event Monitoring 218
9.4.3 Post-event Restoration Problem 219
9.5 Simulation Results 221
9.6 Conclusion 231
References 232
10 Control of Cooperative Unmanned Aerial Vehicles: Review of Applications, Challenges, and Algorithms 234
Abbreviations 234
10.1 Introduction 235
10.2 Applications and Literature Review 236
10.2.1 Search and Rescue 237
10.2.2 Surveillance 238
10.2.3 Localization and Mapping 240
10.2.4 Military Applications 241
10.2.4.1 Reconnaissance Strategy 242
10.2.4.2 Penetrating Strategy 242
10.3 Challenges 243
10.4 Algorithms 245
10.4.1 Consensus Strategies 245
10.4.1.1 Graph Theory Basics in Communication Systems 246
10.4.1.2 Consensus Control Theory 247
10.4.1.3 Consensus Recent Researches 248
10.4.2 Flocking Based Strategies 250
10.4.2.1 Flocking Control Theory 250
10.4.2.2 Flocking Recent Researches 251
10.4.3 Guidance Law Based Cooperative Control 253
10.4.3.1 Guidance Law Based Recent Researches 254
10.5 Summary and Conclusion 255
Bibliography 255
11 An Optimal Approach for Load-Frequency Control of Islanded Microgrids Based on Nonlinear Model 261
Nomenclature 261
11.1 Introduction 262
11.2 Dynamic Model of Microgrid 265
11.3 The Proposed Intelligent Control Method 267
11.4 Simulation and Results 271
11.5 Conclusion 276
References 277
12 Photovoltaic Design for Smart Cities and Demand Forecasting Using a Truncated Conjugate Gradient Algorithm 280
Abbreviations 281
12.1 Introduction 282
12.2 Objectives and Targets 283
12.3 Literature Review 283
12.4 Rule-Based Neural Network Structure 289
12.5 The Proposed Model 290
12.6 Results and Discussion 291
12.7 Conclusion 296
References 297
Index 299

Produkt-Details

• Erscheint lt. Verlag: 22.2.2020
• Reihe/Serie: Advances in Intelligent Systems and Computing
• Zusatzinfo: X, 304 p. 90 illus., 67 illus. in color.
• Sprache: englisch
• Themenwelt: Technik ► Bauwesen
• Themenwelt: Technik ► Elektrotechnik / Energietechnik
• Themenwelt: Technik ► Maschinenbau
• Schlagworte: intelligent transportation • interdependent infrastructure • large-scale complex networks • machine learning • optimization theory • smart cities • Smart Power Grids
• ISBN-10: 3-030-34094-5 / 3030340945
• ISBN-13: 978-3-030-34094-0 / 9783030340940