Invariant set approach to actuator fault tolerant control

Abstract

We present a new actuator fault-tolerant control strategy based on the separation of sets that characterise healthy system operation from sets that characterise faulty operation. The new scheme is an extension of a similar strategy recently proposed by the authors. In the present paper we propose a new criterion for fault detection and isolation that leads to less conservative conditions for fault tolerant closed-loop stability. These new conditions employ discrete-time models for the plant, reference system and observers and allow for quicker fault detection and consequent reconfiguration of the controller.