Cyber-Physical Distributed Systems

Huadong Mo, PhD, is Senior Lecturer in the School of Engineering and Information Technology at the University of New South Wales. He received his doctorate from the City University of Hong Kong in the area of cyber-physical system reliability engineering. Giovanni Sansavini, PhD, is Associate Professor at the Reliability and Risk Engineering Laboratory, Institute of Energy and Process Engineering, ETH Zurich, Switzerland. He is also the director of Reliability and Risk Engineering Laboratory, in the Institute of Energy and Process Engineering, Department of Mechanical and Process Engineering. He received his doctorate in nuclear engineering in 2010 from Politecnico di Milano, Italy, and a doctorate in engineering mechanics from Virginia Tech in Blacksburg in 2010. Min Xie, PhD, is Chair Professor of Industrial Engineering in the Department of Advanced Design and Systems Engineering, at City University of Hong Kong. He received his doctorate in Quality Technology in 1987 from Linkoping University in Sweden and was elected as a Fellow of the IEEE in 2006.

• Preface vList of Acronyms and Abbreviations ix  Introduction 1Challenges of Traditional Physical and Cyber Systems 1Research Trends in Cyber-Physical Systems (CPSs) 3Stability of CPSs 3Reliability of CPSs 6Opportunities for CPS Applications 7Managing Reliability and Feasibility of CPSs 7Ensuring Cybersecurity of CPSs 9  Fundamentals of CPSs 13Models for Exploring CPSs 14Control-Block-Diagram of CPSs 14Control Signal in CPSs 14Degraded Actuator and Sensor 14Time-Varying Model of CPSs 15Implementation in TrueTime Simulator 16Introduction of TrueTime Simulator 16Architecture of CPSs in TrueTime 17Evaluation and Verification of CPSs 18CPS Performance Evaluation 18CPS Performance Index 18Reliability Evaluation of CPSs 19CPS Model Verification 20CPS Performance Improvement 21PSO-Based Reliability Enhancement 22Optimal PID-Automatic Generation Control (AGC) 23  Stability Enhancement of CPSs 29Integration of Physical and Cyber Models 30Basics of Wide-Area Power Systems (WAPS) 30Physical Layer 30Cyber Layer 31WAPS Realized in TrueTime 32An Illustrative WAPS 33Illustrative Physical Layer 33Illustrative Cyber Layer 34Illustrative Integrated System 36Settings of Stability Analysis 36Settings of Delay Predictions 37Settings of Illustrative WAPS 37Cases for Illustrative WAPS 38Hidden Markov Model (HMM)-Based Stability Improvement 38Online Smith Predictor 38Initialization of Discrete HMM (DHMM) 39Parameter Estimation of DHMM 41Delay Prediction via DHMM 43Smith Predictor Structure 44Delay Predictions 44Settings of DHMM 45Prediction Comparison 46Performance of Smith Predictor 47Settings of Smith Predictor 47Analysis of Case 1 47Analysis of Case 2 48Stability Enhancement of Illustrative WAPS 49Eigenvalue Analysis and Delay Impact 49Sensitivity Analysis of Network Parameters 49Optimal AGC 50Optimal Controller Performance 50Scenario 1 Analysis 51Scenario 2 Analysis 51Scenario 3 Analysis 52Scenario 4 Analysis 52Robustness of Optimal AGC 52  Reliability Analysis of CPSs 65Conceptual Distributed Generation Systems (DGSs) 65Mathematical Model of Degraded Network 65Model of Transmission Delay 66Model of Packet Dropout 67Scenarios of Degraded Network 68Modeling and Simulation of DGSs 69DGS Model 69Preliminary Model 69Power Source Model 70Data Interpolation 71Reliability Estimation Via Optimal Power Flow (OPF) 71Data Prediction 71Monte Carlo Simulation (MCS) of DGSs 73OPF of DGSs 74Actual Cost and Reliability Analysis 75OPF of DGSs Against Unreliable Network 76Settings of Networked DGSs 76OPF Under Different Demand Levels 78OPF Under Entire Period 79  Maintenance of Aging CPSs 87Data-driven Degradation Model for CPSs 88Degraded Control System 88Parameter Estimation via EM Algorithm 89Load Frequency Control (LFC) Performance Criteria 90Maintenance Model and Cost Model 91Performance Based Maintenance (PBM) Model 91Cost Model 93Applications to DGSs 94Output of Aging Generators 94Impact of Aging on DGSs 94Settings of Aging DGSs 94Validations of Generator Performance Indexes 95Quantitative Aging Impact 96Applications to Gas Turbine Plant 98Settings of Networked DGS Sensitivity Analysis of PBM 98Impact of Degradation on LFC 98Numerical Sensitivity Analysis 98Pictorial Sensitivity Analysis 99Optimal Maintenance Strategy 100Maintenance Models Comparison 100  Game Theory Based CPS Protection Plan 109Vulnerability Model for CPSs 110Multi-state Attack-Defence Game 111Backgrounds of Game Model for CPSs 111Mathematical Game Model 112Attack Consequence and Optimal Defence 113Damage Cost Model 113Attack Uncertainty 114Optimal Defence Plan 115Applications to DGSs with Uncertain Cyber-Attacks 116Settings of Game Model 116Optimal Protection with Constant Resource Allocation 116Impact Under Constant Case 116Optimal Constant Resource Allocation Fraction 117Optimal Protection with Dynamic Resource Allocation 118Vulnerability Model Under Dynamic Case 119Optimal Dynamic Resource Allocation Fraction 120Optimization Results Justification 121  Bayesian Based Cyberteam Deployment 125Poisson Distribution based Cyber-attacks 125Impacts of DoS Attack 125Poisson Arrival Model Verification 126Average Arrival Attacks 127Cost of Multi-node Bandit Model 128Regret Function of Worst Case 128Upper Bound on Cost 129Thompson-Hedge Algorithm 130Hedge Algorithm 130Details of Thompson-Hedge Algorithm 131Separation of Target Regret 132Upper Bound of Λ_1 133Upper Bound of Λ_2 133Upper Bound of Regret R^TH 134Applications to Smart Grids 135Operation Cost of Smart Grid 135Numerical Analysis of Cost Sequences 137Performance of Thompson-Hedge Algorithm 137Comparison Study Against R.EXP3 137Sensitivity to the Variation 140  Recent Advances in CPS Modeling, Stability and Reliability 145Modeling Techniques for CPS Components 145Inverse Gaussian Process 145Hitting Time to a Curved Boundary 146Estimator Error 147Theoretical Stability Analysis 148Impacts of Uncertainties 148Small Gain Theorem based Stability Criteria 149Robust Stability Criteria 150Game Model for CPSs 151  References 153Index 177