Sequential Multiple-Model Detection of Target Maneuver Termination

摘要

Although methods for target maneuver onset detection (MOD) abound, few have been proposed for target maneuver termination detection (MTD), since MOD is easier and considered more important. Without MTD, however, decision based maneuvering target tracking is not complete and may suffer from performance degradation and computational waste. This paper proposes sequential testing methods for MTD in the multiple model framework. The celebrated CUSUM and SSPRT are adopted to detect maneuver termination, and the likelihood functions involved in these two tests are approximated by the outputs of the interacting multiple model (IMM) estimator, which is one of the most widely used algorithms for maneuvering target tracking. An Illustrative example is provided and the simulation results demonstrate the applicability of our algorithms.