Prebce to the First Edition
Acknowledgments
Preface to the Second Edition
Operators and Notational Conventions
1 Introduction
1.1 Dynamic Systems
1.2 Models
1.3 An Archetypical Problem----ARX Models and the Linear Least Squares Method
1.4 The System Identification Procedure
1.5 Organization of the Book
1.6 Bibliography
part i: systems and models
2 Time-Invariant Linear Systems
2.1 Impulse Responses, Disturbances, and Transfer Functions
2.2 Frequency-Domain Expressions
2.3 Signal Spectra
2.4 Single Realisation Behavior and Ergodicity Results (*)
2.5 Multivariable Systems (*)
2.6 Sununary
2.7 Biblingraphy
2.8 Problems
Appendis 2A: Proof of Theorem 2.2
Appendis 2B: Proof of Theorem 2.3
Appendis 2C: Covariance Formulas
3 Simulation and Prediction
3.1 Simulation
3.2 Prediction
3.3 Observers
3.4 Summary
3.5 Bibliography
3.6 Problems
4 Models of Linear Time-Invariant Systems
4.1 Linear Models and Sets of Linear Models
4.2 A Family of Transfer-Function Models
4.3 State-Space Models
4.4 Distributed Parameter Models (*)
4.5 Model Sets, Model Structures, and Identifiability: Some Formal Aspects(*)
4.6 Identifiability of Some Model Structures
4.7 Summary
4.8 Bibliography
4.9 Problems
Appendix 4A: Identifiability of Black-Box Multivariable Model Structures
5 Models for Time-varying and Nonlinear Systems
5.1 Linear Time-Varying Models
5.2 Models with Nonlinearities
5.3 Nonlinear State-Space Models
5.4 Nonlinear Black-Box Models: Basic Principles
5.5 Nonlinear Black-Box Models: Neural Networks, Wavelets and Classical Models
5.6 Fuzzy Models
5.7 Formal Characterization of Models (*)
5.8 Summary
5.9 Bibliography
5.10 Problems
part ii:methods
6 Nonparametric Time- and Frequency-Domain Methods
6.1 Transient-Response Analysis and Correlation Analysis
6.2 Frequency-Response Analysis
6.3 Fourier Analysis
6.4 Spectral Analysis
6.5 Estimating the Disturbance Spectrum (*)
6.6 Summary
6.7 Bibliography
6.8 Problems
Appendix 6A: Derivation of the AsymPtotic Properties of the Spectral Analysis Estimate
7 Parameter Estimation Methods
7.1 Guiding Principles Behind Parameter Estimation Methods
7.2 Minimising Prediction Errors
7.3 Linear Regressions and the Least-Squares Method
7.4 A Statistical Framework for Parameter Estimation and the Maximum Likelihood Method
7.5 Correlating Prediction Errors with Past Data
7.6 Instrumentatwriable Methods
7.7 Using Frequency Domain Data to Fit Linear Models (*)
7.8 Summary
7.9 Bibliography
7.10 Problems
Appendix 7A: Proof of the Cramer-Rao Inequality
8 Convergence and Consistency
8.1 Introduction
8.2 Conditions on the Data Set
8.3 Prediction-Error Approach
8.4 Consistency and Identifiability
8.5 Linear Time-Invariant Models: A Frequency-Domain Description of the Limit Model
8.6 The Correlation Approach
8.7 Summary
8.8 Bibliography
8.9 Problems
9 Asymptotic Distribution of Parameter Estimates
9.1 Introduction
9.2 The Prediction-Error Approach: Basic Theorem
9.3 Expressions for the Asymptotic Variance
9.4 Frequency-Domain Expressions for the Asymptotic Variance
9.5 The Correlation Approach
9.6 Use and Relevance of Asymptotic Variance Expressions
9.7 Summary
9.8 Bibliography
9.9 Problems
Appendix 9A: Proof of Theorem 9.1
Appendix 9B: The Asymptotic Parameter Variance
10 Computing the Estimate
10.1 Linear Regressions and beast Squares
10.2 Numerical Solution by Iterative Search Methods
10.3 Computing Gradients
10.4 Two-Stage and Multistage Methods
10.5 Local Solutions and Initial Values
10.6 Subspace Methods for Estimating State Space Models
10.7 Summary
10.8 Bibliography
10.9 Problems
11 Recursive Estimation Methods
11.1 Introduction
11.2 The Recursive Least-Squares Algorithm
11.3 The Recursive IV Method
1l.4 Recursive Prediction-Error Methods
11.5 Recursive Pseudolinear Regressions
11.6 The Choice of Updating Step
11.7 Implementation
11.8 Summary
11.9 Bibliography
11.10 Problems
Appendix 11A: Techniques for Asymptotic Analysis of Recursive Algorithms
11A Problems
part iii: user's choices
12 Options and Objectives
12.1 Options
12.2 Objectives
12.3 Bias and Variance
12.4 Summary
12.5 Bibliography
12.6 Problems
13 Experiment Design
13.1 Some General Considerations
13.2 Informative Experiments
13.3 Input Design for Open Loop Experiments
13.4 Identification in Closed Loop: Identifiability
13.5 Approaches to Closed Loop Identification
13.6 Optimal Experiment Design for High-Order Black-Box Models
13.7 Choice of Sampling Interval and Presampling Filters
13.8 Summary
13.9 Bibliography
13.10 Problems
14 Preprocessing Data
14.1 Drifts and Detrending
14.2 Outliers and Missing Data
14.3 Selecting Segments of Data and Merging Experiments
14.4 Prefiltering
14.5 Formal Design of Prefiltering and Input Properties
14.6 Summary
14.7 Bibliography
14.8 Problems
15 Choice of Identification Criterion
15.1 General Aspects
15.2 Choice of Norm: Robustness
15.3 Variance-Optimal Instruments
15.4 Summary
15.5 Bibliography
l5.6 Problems
16 Model Structure Selection and Model Validation
16.1 General Aspects of the Choice of Model Structure
16.2 A Priori Considerations
16.3 Model Structure Selection Based on Preliminary Data Analysis
16.4 Comparing Model Structures
16.5 Model Validation
16.6 Residual Analysis
16.7 Summary
16.8 Bibliography
16.9 Problems
17 System Identification in Practice
17.1 The Tool: Interactive Software
17.2 The Practical Side of System Identification
17.3 Some Applications
17.4 What Does System Identification Have To Offer?
Appendix I Some Concepts From Probability Theory
Appendix II Some Statistical Techniques for Linear Regressions
II.1 Linear Regressions and the Least Squares Estimate
II.2 Statistical Properties of the Least-Squares Estimate
II.3 Some Further Topics in Least-Squares Estimation
II.4 Problems
References
Subject Index
Reference Index
鄂公网安备 42010302001612号 