Adaptive fault tolerant control for actuator bias of nonlinear systems

This paper deals with active fault tolerant controller design against time-varying actuator fault, aiming to improve the robustness, sensitivity of fault detection and the rapidity of whole diagnosis and compensation procedure. A high-gain observer technique is developed to design a residual signal with the estimation error of system states and the derivatives of system output. Then a reconfiguration controller is directly constructed based on the fault information from diagnosis procedure. A qualitative relation is derived to connect the robustness, rapidity and sensitivity of the proposed fault diagnosis scheme with the observe parameters. By prescribing the observer/controller parameters, the set of detectable faults, the time of detection and compensation and the bound of the closed-system signals are quantified. Finally, the theoretical results are illustrated by a simulation example of surface-mounted permanent magnet synchronous motors (PMSM).