Target Maneuver Detection and Estimation
Douglas E. Williams, Bernard Friedland · 1988
A nonlinear target maneuver detection and estimation algorithm, developed using concepts borrowed from failure detection theory, is presented which is based on the assumption that target maneuvers are a piecewise constant process where transitions from one acceleration level to another occur relatively infrequently in time. The development is a two-step procedure using both separated-bias and non-gaussian random transition theory developed by liriedland [ 1][2]. A nonlinear maneuver detection algorithm is designed to signal when a target maneuver has actually occurred, and a linear separate-bias filter, designed under the assumption that the time of the target maneuver is known, is used to estimate the target acceleration vector. Simulation results are presented to demonstrate the efficacy of the procedure.