Control of Nonlinear Systems Using Multiple Model Black-Box Identification
Résumé
The paper is devoted to development of control algorithms for nonlinear parametrically uncertain systems. Original system dynamics is approximated by a set of local NARX models combined by a special mixing rule. Algorithm for local models' parameters estimation and structure adjustment has been developed. Proposed approach allows straightforward designing of the combined feedforward/feedback controller.
Domaines
AutomatiqueOrigine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...