Extended state filter design for general nonlinear uncertain systems

This paper focuses on the filtering problem of general nonlinear uncertain systems. The extended state filter (ESF), featured with timely estimating the uncertainties of the system, is constructed. The stability of ESF is rigorously presented for a general class of systems with nonlinear unknown dynamics, stochastic process and measurement noises. Moreover, the filtering precision can be timely evaluated by the parameter of the filter algorithm. In addition, it is shown that ESF asymptotically tends to the minimum variance filter for linear system with constant disturbance. The simulation studies on the nonlinear spring-mass-damper system also verify this novel filter.