Improved adaptive filter for maneuvering target tracking

A different method from that of Shalom's (1982) variable dimension filter (VDF) method in deciding the initial maneuver moment is presented in this paper. The proposed method is more adaptable to actual cases and presents better tracking property than that of VDF. In order to achieve better real-time computation, the approach of using a Kalman filter steady-gain for target tracking is also proposed. Monte Carlo simulation is applied to the presented approach and significant results have been achieved. >