CONVERGENCE ANALYSIS OF THE EXTENDED KALMAN FILTER USED AS AN OBSERVER FOR NONLINEAR DETERMINISTIC DISCRETE-TIME-SYSTEMS

In this paper, convergence analysis of the extended Kalman filter (EKF ), when used as an observer for nonlinear deterministic discrete-time systems, is presented. Based on a new formulation of the first-order l inearization technique, sufficient conditions to ensure local asymptot ic convergence are established, Furthermore, it is shown that the desi gn of the arbitrary matrix, namely R(k) in the paper, plays an importa nt role in enlarging the domain of attraction and then improving the c onvergence of the modified EKF significantly, The efficiency of this a pproach, compared to the classical version of the EKF, is shown throug h a nonlinear identification problem as well as a state and parameter estimation of nonlinear discrete-time systems.