Identification of time-varying linear models
- 1 January 1983
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 22, 604-606
- https://doi.org/10.1109/cdc.1983.269589
Abstract
This paper presents a convergence analysis of a least squares type recursive identification algorithm for time-varying linear-in-the parameters models. It is shown that under persistent excitation conditions, the number of the associated gain matrix is bounded. This is obtained by normalizing the observation vector entering in the identification algorithm. This allows a simple convergence analysis to study the influence of plant parameters changes and noise in the estimates.Keywords
This publication has 5 references indexed in Scilit:
- Persistence of excitation conditions and the convergence of adaptive schemesIEEE Transactions on Information Theory, 1984
- Convergence analysis of recursive identification algorithms with forgetting factorAutomatica, 1983
- Persistence of excitation in extended least squaresIEEE Transactions on Automatic Control, 1983
- Exponential convergence of adaptive identification and control algorithmsAutomatica, 1982
- Unification of discrete time explicit model reference adaptive control designsAutomatica, 1981