An adaptive kalman filter with dynamic resealing of process noise
Zhansheng Duan, Chongzhao Han, Haodong Dang · 2003
Abstract In this paper, a novel adaptive Kalman filter is proposed to track maneuvering target with unknown process noise statistics. The process noise statistics are estimated by Sage’s algorithm firstly; then a residual based hypothesis testing method is used to detect the likely divergence of the filter; once detected, a timevarying scale factor matrix is proposed to modify the estimated value of the process noise covariance, thus an adaptive Kalman filtering algorithm with dynamic rescaling of the process noise is derived. By using the mean-adaptive acceleration model, validity of the proposed algorithm is verified by means of Monte-Carlo simulations in two typical target maneuver scenarios.