内容简介:
This monograph focuses on characterizing the stability and performance consequences of inserting limited-capacity communication networks within a control loop. The text shows how integration of the ideas of control and estimation with those of communication and information theory can be used to provide important insights concerning several fundamental problems such as: minimum data rate for stabilization of linear systems over noisy channels; minimum network requirement for stabilization of linear systems over fading channels; and stability of Kalman filtering with intermittent observations.
A fundamental link is revealed between the topological entropy of linear dynamical systems and the capacities of communication channels. The design of a logarithmic quantizer for the stabilization of linear systems under various network environments is also extensively discussed and solutions to many problems of Kalman filtering with intermittent observations are demonstrated.
Analysis and Design of Networked Control Systems will interest control theorists and engineers working with networked systems and may also be used as a resource for graduate students with backgrounds in applied mathematics, communications or control who are studying such systems.
英文目录:
Preface
1 Overview of Networked Control Systems
1.1 Introduction and Motivation
1.1.1 Components of NCS
1.1.2 Brief History of NCS
1.1.3 Challenges in NCS
1.2 Preview of the Book
References
2 Entropies and Capacities in Networked Control Systems
2.1 Entropies
2.1.1 Entropy in Information Theory
2.1.2 Topological Entropy in Feedback Theory
2.2 Channel Capacities
2.2.1 Noiseless Channels
2.2.2 Noisy Channels
2.3 Control Over Communication Networks
2.3.1 Quantized Control Over Noiseless Networks
2.3.2 Quantized Control Over Noisy Networks
2.4 Estimation Over Communication Networks
2.4.1 Quantized Estimation Over Noiseless Networks
2.4.2 Data-Driven Communication for Estimation
2.4.3 Estimation Over Noisy Networks
2.5 Open Problems
References
3 Data Rate Theorem for Stabilization Over Noiseless Channels
3.1 Problem Statement
3.2 Classical Approach for Quantized Control
3.3 Data Rate Theorem for Stabilization
3.3.1 Proof of Necessity
3.3.2 Proof of Sufficiency
3.4 Summary
References
4 Data Rate Theorem for Stabilization Over Erasure Channels
4.1 Problem Formulation
4.2 Single Input Case
4.2.1 Proof of Necessity
4.2.2 Proof of Sufficiency
4.3 Multiple Input Case
4.4 Summary
References
5 Data Rate Theorem for Stabilization Over Gilbert-Elliott Channels
5.1 Problem Formulation
5.2 Preliminaries
5.2.1 Random Down Sampling
5.2.2 Statistical Properties of Sojourn Times
5.3 Scalar Systems
5.3.1 Noise Free Systems with Bounded Initial Support
5.3.2 Proof of Necessity
5.3.3 Proof of Sufficiency
5.4 General Stochastic Scalar Systems
5.4.1 Proof of Necessity
5.4.2 Proof of Sufficiency
5.5 Vector Systems
5.5.1 Real Jordan Form
5.5.2 Necessity
5.5.3 Sufficiency
5.5.4 An Example
5.6 Summary
References
6 Stabilization of Linear Systems Over Fading Channels
6.1 Problem Formulation
6.2 State Feedback Case
6.2.1 Parallel Transmission Strategy
6.2.2 Serial Transmission Strategy
6.3 Output Feedback Case
6.3.1 SISO Plants
6.3.2 Triangularly Decoupled Plants
6.4 Extension and Application
6.4.1 Stabilization Over Output Fading Channels
6.4.2 Stabilization of a Finite Platoon
6.5 Channel Processing and Channel Feedback
6.6 Power Constraint
6.6.1 Feedback Stabilization
6.6.2 Performance Design
6.6.3 Numerical Example
6.7 Summary
References
7 Stabilization of Linear Systems via Infinite-Level Logarithmic Quantization
7.1 State Feedback Case
7.1.1 Logarithmic Quantization
7.1.2 Sector Bound Approach
7.2 Output Feedback Case
7.2.1 Quantized Control
7.2.2 Quantized Measurements
7.3 Stabilization of MIMO Systems
7.3.1 Quantized Control
7.3.2 Quantized Measurements
7.4 Quantized Quadratic Performance Control
7.5 Quantized H1 Control
7.6 Summary
References
8 Stabilization of Linear Systems via Finite-Level Logarithmic Quantization
8.1 Quadratic Stabilization via Finite-level Quantization
8.1.1 Finite-level Quantizer
8.1.2 Number of Quantization Levels
8.1.3 Robustness Against Additive Noises
8.1.4 Illustrative Examples
8.2 Attainability of the Minimum Data Rate for Stabilization
8.2.1 Problem Simplification
8.2.2 Network Configuration
8.2.3 Quantized Control Feedback
8.2.4 Quantized State Feedback
8.3 Summary
References
9 Stabilization of Markov Jump Linear Systems via Logarithmic Quantization
9.1 State Feedback Case
9.1.1 Feedback Stabilization
9.1.2 Special Schemes
9.1.3 Mode Estimation
9.2 Stabilization Over Lossy Channels
9.2.1 Binary Dropouts Model
9.2.2 Bounded Dropouts Model
9.2.3 Extension to Output Feedback
9.3 Summary
References
10 Kalman Filtering with Quantized Innovations
10.1 Problem Formulation
10.2 Quantized Innovations Kalman Filter
10.2.1 Multi-level Quantized Filtering
10.2.2 Optimal Quantization Thresholds
10.2.3 Convergence Analysis
10.3 Robust Quantization
10.4 A Numerical Example
10.5 Summary
References
11 LQG Control with Quantized Innovation Kalman Filter
11.1 Problem Formulation
11.2 Separation Principle
11.3 State Estimator Design
11.4 Controller Design
11.5 An Illustrative Example
11.6 Summary
References
12 Kalman Filtering with Faded Measurements
12.1 Problem Formulation
12.2 Stability Analysis of Kalman Filter with Fading
12.2.1 Preliminaries
12.2.2 Mean Covariance Stability
12.3 A Numerical Example
12.4 Summary
References
13 Kalman Filtering with Packet Losses
13.1 Networked Estimation
13.1.1 Intermittent Kalman Filter
13.1.2 Stability Notions
13.2 Equivalence of the Two Stability Notions
13.3 Second-Order Systems
13.4 Higher-Order Systems
13.4.1 Non-degenerate Systems
13.5 Illustrative Examples
13.6 Proofs
13.6.1 Proof of Theorem 13.3
13.6.2 Proof of Theorem 13.4
13.6.3 Proofs of Results in Sect. 13.4
13.7 Summary
References
14 Kalman Filtering with Scheduled Measurements
14.1 Networked Estimation
14.1.1 Scheduling Problems
14.2 Controllable Scheduler
14.2.1 An Approximate MMSE Estimator
14.2.2 An Illustrative Example
14.2.3 Stability Analysis
14.3 Uncontrollable Scheduler
14.3.1 Intermittent Kalman Filter
14.3.2 Second-Order System
14.3.3 Higher-Order System
14.4 Summary
References
15 Parameter Estimation with Scheduled Measurements
15.1 Innovation Based Scheduler
15.2 Maximum Likelihood Estimation
15.2.1 ML Estimator
15.2.2 Estimation Performance
15.2.3 Optimal Scheduler
15.3 Naive Estimation
15.4 Iterative ML Estimation
15.4.1 Adaptive Scheduler
15.5 Proof of Theorem 15.1
15.6 EM-Based Estimation
15.6.1 Design of b yk
15.7 Numerical Example
15.8 Summary
References
Appendix A: On Matrices
Index
A Background Materials
A.1 Martingales
A.2 Markov Chains
A.3 Weak Convergence
A.4 Miscellany
References
Index