Abstract:

To solve the problems of low precision and divergence of filter caused by unknown system noise statistics in target tracking, a new adaptive square cubature Kalman filter (CKF) algorithm is proposed. A noise statistics estimator designed for nonlinear systems is derived by applying the cubature rule based on the Sage-Husa algorithm. Simulation results show that as compared with the standard square CKF algorithm, the proposed algorithm provides higher accuracy when the system noise statistics is unknown or time-varying.