Tracking of Multiple Maneuvering Targets in Clutter by Joint Probabilistic Data and Maneuver Association

Abstract

A feasible and effective method of tracking multiple maneuvering targets in clutter is introduced. The maneuver is modeled by a Markovian selection from a finite set of acceleration inputs in the target state equation. An extension of the joint probabilistic data association (JPDA) algorithm to the case of maneuvering targets is then presented which employs a joint hypothesis space for the maneuver and measurement-to-track association. The proposed joint probabilistic data and maneuver association (JPDMA) algorithm requires approximately the same amount of computation as the ordinary JPDA algorithm in spite of its additional ability to track maneuvering targets. Computer simulations verify the ability of the JPDMA algorithm to track several maneuvering targets in the presence of sparse clutter.