A Strong Tracking Square Root CKF Algorithm Based on Multiple Fading Factors for Target Tracking

The performance of Cubature Kalman Filter (CKF) will degrade seriously when the theoretical model and real model are not matched due to change of target motion. To solve this problem, a new strong tracking square root CKF algorithm is proposed in this paper. Suboptimal multiple fading factors are used in the proposed method, which can adjust the structural parameters of the filter, and improve the performance of target state tracking. Different from the CKF method using single fading factor, channels are faded with multiple parameters in the proposed method, and weighted square root is also used to avoid the asymmetry in the calculation of predicted covariance matrix. Simulation results show that as compared with CKF method using single fading factor, the proposed method provides higher accuracy.