Learning Algorithm for Tracking Hypersonic Targets in Near Space

Abstract

With the development of hypersonic vehicles in near space such as X-51A, HTV-2 and so on, tracking for them is becoming a new task and hotspot. In this paper, a learning tracking algorithm is introduced for hypersonic targets, especially for the sliding jump maneuver. Firstly the algorithm uses the Sine model, which makes the tracking model more close to the particular maneuver, next two Sine models different in angular velocity are used into IMM algorithm, and it learns the target tracking error characteristics to adjust the sampling rate adaptively. The algorithm is compared with the single accurate model algorithm and general IMM algorithms with fixed sampling rate. Through simulation experiments it is proved that the algorithm in this paper can improve the tracking accuracy effectively.