Adaptive tracking of abruptly maneuvering targets

An adaptive estimator solution to a non-linear state estimation problem is presented. The specific problem being addressed is that of tracking a maneuvering target in three-dimensional space by spherical observations (radar data). Two techniques are presented, both of which model target maneuvers as a semi-Markov process. This paper describes the two techniques and presents computer simulation results. This work was supported by NSWC Dahlgren Lab., project no. 382922-1.