Spectral analysis in linear process identification—a new look at an old problem
作者:
Y. C. ZHU,
M. H. KLOMPSTRA,
期刊:
International Journal of Control
(Taylor Available online 1991)
卷期:
Volume 53,
issue 6
页码: 1449-1465
ISSN:0020-7179
年代: 1991
DOI:10.1080/00207179108953685
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
The problem of estimating the transfer function (frequency response) of a linear process by means of a modified spectral analysis method is treated. A method of determining the maximum time lags of the correlation functions which makes use of the availablea prioriphysical knowledge about the process and the input signal is proposed. This leads, in a natural way, to the use of anasymmetricalwindow technique. Some simulation results are used to illustrate the method. Based on the asymptotic properties of the estimate, the input design is studied. Finally it will be shown that this work can be used for the estimation of an upper bound of the model uncertainty in the frequency domain.
点击下载:
PDF (464KB)
返 回