Efficient Multiple Model Adaptive Estimation in Ballistic Missile Interception Scenarios

Abstract

A novel efe cient algorithm, featuring a highly reduced computational load, is presented for multiple model adaptive estimation in a future real-life ballistic missile defense scenario, where the blind incoming target (having no information on the interceptor’ s state ) performs a bang ‐bang evasive maneuver characterized by a random switching time. The efe ciency of the algorithm derives mainly from its exploitation of the special structure of the hypothesis space in this problem to drastically reduce the number of concurrently active e lters in the bank without incurring any signie cant performance degradation. The proposed algorithm’ s efe ciency allows a substantial increase in the resolution of the discretized hypothesis space, thus enhancing considerably the attainable estimation performance. The effect of the new estimator’ s performance on guidance accuracy is examined. The homing performance of various perfect information guidance laws using this efe cient estimation method is compared, via Monte Carlo simulations, to the use of a Kalman e lter incorporating a shaping e lter representing the random target maneuver. The results demonstrate the superiority and viability of the proposed method.