A multiple model adaptive architecture for the state estimation in discrete-time uncertain LPV systems

Abstract

This paper addresses the problem of multiple model adaptive estimation (MMAE) for discrete-time linear parameter varying (LPV) systems that are affected by parametric uncertainty. The MMAE system relies on a finite number of local observers, each designed using a selected model (SM) from the set of possible plant models. Each local observer is an LPV Kalman filter, obtained as a linear combination of linear time invariant (LTI) Kalman filters. It is shown that if some suitable distinguishability conditions are fulfilled, the MMAE will identify the SM corresponding to the local observer with smallest output prediction error energy. The convergence of the unknown parameter estimation, and its relation with the varying parameters, are discussed. Simulation results illustrate the application of the proposed method.