System identification theory and application

《系统辨识理论及应用英文》

chapter 1 introduction
1.1 classification of mathematic models of system and modelling methods
1.1.1 signification of model

1.1.2 representation forms of models
1.1.3 classification of mathematic models

1.1.4 basic methods to establish mathematic model
1.1.5 basic principles followed for modeling

1.2 definition, content and procedure of identification

1.2.1 definition of identification

1.2.2 content and procedure of identification

1.3 error criteria usually used in identification

1.3.1 output error criterion
1.3.e input error criterion
1.3.3 generalized error criterion
1.4 classification of system identification

1.4.1 off-line identification
1.4.2 on-line identification
problems
chapter 2 commonly used input signals for system identification
.2.1 selective criteria of input signal for system identification

2.2 white noises and its generating methods
2.2.1 white noise process

2.2.2 white noise sequence
2.2.3 generating methods of white noise sequence
2.3 generation of pseudorandom binary sequence-m-sequence and its properties

2.3.1 pseudorandom noise
2.3.2 generating method of m-sequence
2.3.3 properties o{ m-sequence
2.3.4 autocorrelation function of two-level m-sequence

2.3.5 power spectral density of two-level m-sequence

problems

chapter 3 classical identification methods of linear system
3.1 identify impulse response of linear system by use of m-sequence
3.2 obtain transfer function from impulse function

3.2.1 transfer function g(s) of continuous system
3.2.2 transfer function of discrete system—impulse transfer function g(z-1)

problems
chapter 4 canonical expression of dynamic systems
4.1 parsimony principle

4.2 representations of difference equation and state equation of linear system
4.2.1 representation of difference equation of linear time-invariant system
4.2.2 representation of state equation of linear system
4.3 deterministic canonical state equations

4.3.1 controllable form of canonical state equation ⅰ

4.3.2 controllable form of canonical state equation ⅱ

4.3.3 observable form of canonical state equation ⅰ

4.3.4 observable form of canonical state equation ⅱ

4.3.5 observable form of canonical state equation ⅰ of mimo system
4.3.6 observable form of canonical state equation ⅱ of mimo system
4.4 deterministic canonical difference equations

4.5 stochastic canonical state equations
4.6 stochastic canonical difference equations
4.7 prediction error equation

problems
chapter 5 least-squares identification

5.1 least square method
5.1.1 algorithns of least-square estimation

5.1.2 input signals for least-squares estimation

5.1.3 probability properties of least-squares estimation
5.2 a kind of least squres which need not invert matrix
5.3 recursive least squares

5.4 auxiliary variable method
5.5 recursive auxiliary variable method
5.6 generalized least squares
5.7 an alternative generalized least squares technique (hsia method)

5.8 extended matrix method
5.9 multistage least squares
5.9.1 the first algorithm
5.9.2 the second algorithm

5.9.3 the third algorithm

5.10 fast multistage least squares

problems

chapter 6 maximum-likelihood identification
6.1 maximum-likelihood method for parameter identification

6.1.1 principle of maximum likelihood

6.1.2 maximum-likelihood estimation of system parameters
6.2 recursive maximum-likelihood method
6.2.1 approximate recursive maximum-likelihood method
6.2.2 recursive newton-raphson maximum-likelihood algorithm
6.3 achievable precision of parameter estimations

problems

chapter 7 identification methods of time-varying parameters

7.1 forgetting factor method, rectangular window method and kalman filter method
7.1.1 forgetting factor method

7.1.2 rectangular window method
7.1.3 kalman filter method

7.2 an identification method of time-varying parameters with automatically adjusted forgetting factor
7.3 an identification method using broken-line segments to approximate to time-varying parameter

problems

chapter 8 identification of multi-input multi-output systems
8.1 least-squares identification of multi-input multi-output systems
8.2 maximum-likelihood identification of multi-input multi-output system: relaxation algorithm

8.3 use square-wave pulse function to identify state equation of linear time-varying system

8.3.1 expansion of state station in square-wave pulse series
8.3.2 identification of matrix a(t)

8.3.3 identification of matrix b(t)

8.4 identification of multi-input multi-output linear time-varying system

by use of piecewise multiple chebysheve polynomials
8.4.1 definition of piecewise multiple chebyshev polynomials and their main properties
8.4.2 parameter identification of multi-input multi-output linear time-varying system by use of piecewise multiple chebyshev polynomials

problems
chapter 9 some other kinds of identification methods

9.1 a simple recursive algorithm——method of stochastic approximation

9.1.1 basic principle of stochastic approximation

9.1.2 method of parameter estimation by means of approximation

9.1.3 random newton method
9.2 two kinds of recursive least squares based on different concepts

9.2.1 least-squares estimation recuring with number of observation equations
9.2.2 recursive least-squares estimation varying with number of unknown parameters
9.2.3 error model best estimation of trajectory derived based on recursive least squares
9.3 recursive generalized extended least squares for identification of box-jenkins model
9.4 innovations-modified least squares for identification of box-jenkins model
9.4.1 increasing-prameter recursive formulas of least squares
9.4.2 identification of car(p) model
9.4.3 elimination of derivations and determination of order for ma part

problems

chapter 10 establishment of random time series models

10.1 regressive model
10.1.1 first-order linear regressive model
10.1.2 polynomial regressive model

10.2 autoregressive model of stationary time series

10.3 moving average models of stationary time series
10.4 autoregressive moving ,average model of stationary time series
10, 5 model of non-stationary time series
problems

chapter 11 structure identification of system
11.1 order estimation of models
11.1.1 determine order according to variance of residual errors

11.1.2 akaike information criterion for determining order

11.1.3 determine order by use of white residual errors

11.1.4 use zero-pole 'cancellation to check order
11.1.5 determine order by use of ratio between determinants

11.1.6 determine order by use of hankel matrix
11.2 a non-recursive algorithm to identify order and parameters of model at the same time

11.3 a recursive algorithm to identify order and parameters of model at the same time

11.4 structure identification of multivariable carma model

11.4.1 recursive least-squares estimation of parameters

11.4.2 determine the order of submodel
11.4.3 determinations of succinct parameter model, suborder and time-delay

problems
chapter 12 identification of closed-loop system
12.1 discrimination methods of closed-loop systems

12.1.1 method of spectral factor decomposition
12.1.2 method of likelihood ratio test

12.2 identifiable concept of closed-loop system
12.3 identification of single-input single-output closed-loop system
12.3.1 direct identification

12.3.2 indirect identification
12.4 identification of multi-input multi-output closed-loop system

12.4.1 autoregressive model identification method

12.4.2 identification method changing feedback matrix

problems
chapter 13 application of system identification to parameter identification of aircraft
13.1 foreword

13.1.1 identification of aerodynamic parameters
13.1.2 identification of aerothermodynamic parameters

13.1.3 parameter identification of structural dynamics

13.1.4 identification of modal parameters for rock of liquid

13.1.5 error coefficient identification of inertial instrument

13.2 maximum likelihood identification of noise model for target seeker of missile

13.2.1 description of seeker noise model

13.2.2 maximum likelihood identification of noise model parameters

13.2.3 identification of noise model for seeker-target line-of-sight angular velocity
13.2.4 identification of noise model for target approaching velocity

13.2.5 check of noise model

13.2.6 an example of maximum likelihood identification
13.3 modelling of output noise for seeker system by use of time series

13.3.1 design of scheme
13.3.2 establishment of noise model

13.3.3 parameter identification of noise model

13.3.4 example of identification by use of time series method
13.4 application of system identification to pneumatic parameter identification of aircraft

13.4.1 identification of aerodynamic parameters for tactical missile

13.4.2 examples of closed-loop identification
chapter 14 applicatiom of neural network to system identification

14.1 brief introduction of neutral network
14.1.1 development survey of neutral network

14.1.2 structure and type of neutral network
14.2 identification of linear system

14.2.1 identification of linear system based on single-layer neutral network
14.2.2 identification methods of linear system based on single-layer adaline network
14.3 application of bp algorithm to neutral network

14.3.1 brief introduction of bp network

14.3.2 mathematical principle of bp network

14.4 identification of linear time-varying system
14.4.1 structure of network and analysis of its approximating ability
14.4.2 learning algorithm

14.4.3 results of simulation
references