Variable structure robust state and parameter estimator

Citation
As. Poznyak et Jc. Martinez, Variable structure robust state and parameter estimator, INT J ADAPT, 15(2), 2001, pp. 179-208
Citations number
28
Categorie Soggetti
AI Robotics and Automatic Control
Journal title
INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING
ISSN journal
08906327 → ACNP
Volume
15
Issue
2
Year of publication
2001
Pages
179 - 208
Database
ISI
SICI code
0890-6327(200103)15:2<179:VSRSAP>2.0.ZU;2-2
Abstract
The problem of simultaneous robust state and parameters estimation for a cl ass of SISO non-linear systems under mixed uncertainties (unmodelled dynami cs as well as observation noises) is addressed. A non-linear variable structure robust 'observer-identifier' is introduced to obtain the corresponding estimates. Lie derivative technique is used to obtain the observability conditions for the equivalent extended non-linear system. It is shown that, in general, the extended system can lose the glob al observability property and a special procedure is needed to work well in this situation. The suggested adaptive observer has the non-linear high-ga in observer structure with adjusted parameters that provides 'a good' upper bound for the identification error performance index. The van der Monde tr ansformation is used to derive this bound which turns out to be tight. Thre e examples dealing with a simple pendulum, the Duffing equation and the van del Pol oscillator are considered to illustrate the effectiveness of the s uggested approach. Copyright (C) 2001 John Wiley & Sons, Ltd.