An Interacting Multiple Model Algorithm Based On Cubature Kalman Filter
Biao Zhang, Yachun Wan, Chaojun Sun, Qiang Yao, Gang Fang · 2023
The interacting multiple model(IMM) based on different motion models can mix the target state estimation,and be proved to be the most effective target state estimator.Based on the interacting multiple model composed of the constant velocity (CV) model,the adaptive coordinated turn(ACT) model and the current statistical(CS) model,this paper adopts the cubature kalman filter(CKF) to realize the nonlinear filter. The simulation of a maneuvering target tracking shows that the algorithm can estimate the rate of turn in real time,has the ability to recognize maneuvers, and has a high tarcking accuracy.