Quantitative detection for actuator faults of nonlinear systems

This paper studies the problem of fault detection in a class of nonlinear systems with time-varying actuator fault. A nonlinear extended observer is developed to estimate system states as well as the derivatives of system output, then a fault diagnosis (FD) scheme is presented. The robustness, rapidity and sensitivity of the proposed FD method are rigorously analyzed. By preselecting the observer parameters, the set of detectable faults, the time of detection and the bound of the closed-system signals are strictly quantified. Finally, the theoretical results are illustrated by a simulation example of surface-mounted permanent magnet synchronous motors (PMSM).