STABILIZING CONTROLLER DESIGN FOR LINEAR UNKNOWN UNSTABLE SYSTEMS: SUCCESSIVE LEARNING IDENTIFICATION AND MAXIMALLY ROBUST CONTROLLER DESIGN

Abstract

Abstract For an unknown, unstable linear time invariant (LTI) plant with output disturbances, an effective scheme for finding a stabilizing controller is proposed in terms of successive data-based plant identification and model-based controller design. The identification is performed by using an optimal parameter selection technique, and the controller design is based on a robust controller design technique. The main feature of this scheme is that there is almost no requirement of a priori information about the plant. Extensive simulation studies confirm that the scheme is both reliable and efficient. For illustration, a numerical example is presented.