Adaptive /spl alpha/-/spl beta/ tracking filter using adjustment of process noise

An adaptive alpha-beta tracking filter is proposed for tracking maneuvering targets with a track-while-scan system. The tracking filter gain is updated on-line corresponding to the adjusted process noise variance which is obtained via time averaging of the process over a sliding window. The adjusted process noise variance is used to compute the tracking index for each epoch. It is shown via simulation that the proposed approach provides robust and accurate position estimates during the target maneuver while the performances of the conventional Kalman filter and the conventional alpha-beta tracker are shown much degraded