Robust identification of Linear Parameter Varying systems

Abstract

In this paper we study the robust identification problem for a special class of Linear Parameter Varying (LPV) systems, when it is assumed that there is only one varying parameter and all the states of the system under identification can be measured. We consider here a worst-case approach, i.e., we look for a set of models in a class of systems defined a priori, which could have generated the available bounded-error data. The proposed method performs the identification recursively and in real time.