Identification of Linear Systems: A Practical Guideline to Accurate Modeling

by J. Schoukens, R. Pintelon

Overview

This book concentrates on the problem of accurate modeling of linear systems. It presents a thorough description of a method of modeling a linear dynamic invariant system by its transfer function. The first two chapters provide a general introduction and review for those readers who are unfamiliar with identification theory so that they have a sufficient background knowledge for understanding the methods described later. The main body of the book looks at the basic method used by the authors to estimate the parameter of the transfer function, how it is possible to optimize the excitation signals. Further chapters extend the estimation method proposed. Applications are then discussed and the book concludes with practical guidelines which illustrate the method and offer some rules-of-thumb.

Product Details

ISBN-13:	9780080407340
Publisher:	Elsevier Science
Publication date:	05/26/1991
Edition description:	1st ed
Pages:	356
Product dimensions:	0.88(w) x 6.14(h) x 9.21(d)

Table of Contents

Chapter headings and selected topics: Preface. A General Introduction to Parameter Estimation. Steps in the identification process. Parameter estimation, an example: measurement of a resistor. The ideal estimator. A Review of Estimation Methods and their Applications. Motives in focusing on the least squares technique. Parametric models. Time domain versus frequency domain. Errors due to noise on the independent variables. A Maximum Likelihood Estimator for Linear Time Invariant Systems. Measurement of a resistance. Estimation of transfer functions: a practical approach. Estimation of transfer functions: a theoretical approach. Numerical considerations. Extensions of the model. Application of ELiS to experimental data. Conclusions. Design of Excitation Signals. Optimization of the time domain behaviour of excitation signals. Optimizing the frequency domain behaviour of excitation signals: design of optimized power spectra. Model Selection. Verification of model validity. Introduction of the model complexity in the cost function. Study of the influence of model errors on the behaviour of the cost function. Optimal experiment strategy. Examples. Estimation of Linear Time Invariant Systems with Delay. The estimation algorithm and its properties. Simulations. Phase Correction of Linear Time Invariant Systems with Digital Allpass Filters. Phase distortion. Optimization strategy. Noise sensitivity. Application of ELiS to Measurement Problems. Modal analysis. Flight flutter data analysis. A Guideline for Transfer Function Estimation. Accurate modeling of a linear analog system. References. Author index. Subject index.

Editorial Reviews

For researchers and practicing engineers faced with constructing models of linear dynamic systems from noisy data, describes a method of modeling a linear dynamic time invariant system by its transfer system. The technique is kept general to allow its application in a wide range of fields. Assumes a basic knowledge of system theory, spectral analysis, discrete time systems, sampled signals, probability theory, elementary statistics, and stochastics. Annotation c. Book News, Inc., Portland, OR (booknews.com)

Booknews