Performance evaluation of models, identified by the least squares method
- 123 Downloads
The least squares method is widely used for the identification of linear, dynamical systems. Here we investigate how the variances of the estimated models affect their performances in various applications, such as transfer function accuracy, prediction, minimum variance control and pole placement servo control design. Some newly developed asymptotic expressions for the variances are applied. Also, some consequences for the choice of input spectrum and model orders are drawn.
KeywordsClosed Loop System Model Order Optimal Input Transfer Function Model Pole Assignment
Unable to display preview. Download preview PDF.
- H. Akaike (1970):"Statistical Predictor Identification". Ann. Inst. Statist. Math. Vol. 22 pp 202–217.Google Scholar
- K. J. Åström (1970):"Introduction to Stochastic Control", Academic Press, N.Y.Google Scholar
- L. Ljung (1984a):"Asymptotic Properties of the Least Squares Method for Estimating Transfer Functions and Disturbance Spectra". Report, Dept. of Electrical Engineering, Linköping University, Linköping, Sweden.Google Scholar
- L. Ljung (1984b):"Asymptotic variance expressions for identified black-box transfer function models". Report, Dept. of Electrical Engineering, Linköping University, Linköping Sweden.Google Scholar
- L. Ljung and Z. D. Yuan (1983):"Properties on Non-Parametric Time-Domain Methods for Estimating Transfer Functions", Report, LiTH-ISY-I-570, Linköping University, Linköping, Sweden.Google Scholar
- Z. D. Yuan and L. Ljung (1983):"Unprejudiced Optimal Input Design for Identification of Transfer Functions". Report, LiTH-ISY-I-0622, Linköping University, Linköping, Sweden.Google Scholar