Fault Detection of Singular Stochastic Systems via Observers

This paper deals with the problem of fault detection (FD) for a class of singular stochastic systems. The correlative FD problem is transformed into the nonlinear dynamic weight system based on probability density function (PDF) modeled by square root B-spline expansions technique. Nonlinear observer is designed to estimate the state. Synchronously, an interrelated Lyapunov function is presented to verify the stability of the error estimation system. The design can detected the fault through an adaptive threshold. Finally, a numerical example is exploited to demonstrate the effectiveness of the compensator obtained.