Sliding mode switching observer-based actuator fault detection and isolation for a class of uncertain systems

Abstract

This paper proposes a fault detection and isolation (FDI) method for a class of uncertain systems with actuator faults. First, based on the multiple model schemes, a bank of parallel sliding mode switching observers are designed to match the fault-free, stuck and loss of effectiveness fault models. A switching strategy is introduced to maintain a sliding motion even in the presence of actuator faults. Stability analysis of the sliding motions and the reachability conditions are given in various fault models. Without the assumption that all states are available, the residual signals are constructed by using the measurable output estimation errors. Then, the fault detection scheme is designed by comparing the residual evaluation function with the threshold. The fault isolation scheme is derived by looking for the minimum residual evaluation function, which is generated by the observer that matches the current system model well. Finally, an L-1011 aircraft model is presented to illustrate the effectiveness of the FDI method obtained in this paper.