In this paper, we present a fuzzy set membership (FSM) filter for state estimation of nonlinear discrete-time systems with unknown but bounded disturbances. First, the Takagi-Sugeno (TS) fuzzy model is used to represent the nonlinear systems. Based on this model, FSM filter is proposed to further improve the numerical accuracy and stability of the nonlinear system estimation. In comparison with the existing fuzzy Kalman (FK) filter, the proposed FSM filter using the efficient and stable updating recursions provides much more accurate estimation results. Simulation results show that the propose...