Reconfigurable-Based Fault-Tolerant Control for Continuous-Time Markov Jump Piecewise-Affine Systems with Allowable Switching Paths

Abstract

In this paper, the reconfigurable control design problem is studied for a class of continuous-time Markov jump piecewise-affine (PWA) systems with sensor and actuator faults. A reconfiguration block is designed to recover mean-square input-to-state stability (ISS) of the faulty closed-loop system under virtual sensor and virtual actuator. Then, a novel algorithm for computing all the allowable switching paths (ASPs) is developed to overcome the issues of redundant empty paths and mismatch regions, and a S-procedure with the ellipsoidal outer approximation estimation is employed to cope with the affine term. Furthermore, a nonlinear approximation is introduced using the PWA technique to approximate the faulty Markov jump nonlinear systems. Finally, one example is illustrated to demonstrate the effectiveness and advantages of the designed reconfigurable control strategy with an ASPs algorithm.