Maneuver Detector for Active Tracking Update Rate Adaptation

The update rate criterion used by Blackman and Van Keuk in [1] is designed to adapt the update rate of the radar when the estimation filter is reliable, which excludes the cases of high maneuvers. In this paper, we thus propose a method to detect the maneuvers, so that urgent pointings can be ordered with a wider beamwidth as soon as a maneuver is detected, to prevent target loss. This is a complementary approach to the Blackman and Van Keuk solution. The target model we opt for is based on the assumption of piecewise constant velocity’s norm, curvature, and torsion, with jumps of those latter parameters at unknown random times. To perform maneuver detection, we use the Variable Rate Particle Filter (VRPF), which is a particle filter that allows considering dynamical systems that jump at unknown random times which differ from the measurement times, and is here fed with the IEKF outputs and used as a maneuver detector.