Prediction and estimation to the maneuvering target in opto-electronic tracking system based on robust extend Kalman filter

With the improvement of target maneuvering performance, noises statistical character changes correspondingly in the opto-electronic tracking system. To solve this problem, robust extend Kalman filter (REKF) based on the acceleration non-zero mean value time-correlation model was given in this paper. Result of the simulation shows that, comparing with extend Kalman filter, the prediction precision of REKF is almost twice as much as that of EKF. And at the same time, REKF has good robustness to overcome the effect coursed by stochastic measurement noises.