Robust State and Fault Estimation for Nonlinear Descriptor Stochastic Systems with Unknown Disturbances

This paper present a nonlinear robust multi-step delayed state and the fault estimation for a class of nonlinear descriptor discrete-time stochastic systems in light of the unknown input filtering framework. Based on the one-step delay fault estimation approach, a novel nonlinear robust multi-step delayed state and fault estimator (NRMSDSFE) is proposed. In order to achieve the aim, the nonlinear descriptor stochastic system with fault and unknown disturbances is first transformed into an equivalent augmented system. Then, it is shown that the previously proposed robust two stage Kalman filter can be applied to yield an optimal robust state and fault estimation and can practically better depict the size and shape of the faults. Simulation results for a single-link robotic manipulator actuated by a brushed DC motor with a nonrigid joint show the effectiveness of the proposed method.