Abstract: To solve the tracking problem of hypersonic glide vehicle (HGV) during maneuvering flight, a robust extended Kalman filter based on maneuvering observer compensation is proposed. Aiming at the problem that the traditional Kalman filter cannot track stably due to model error, firstly, the HGV dynamic model and the space based infrared measurement model are established. Secondly, a maneuvering observer is designed to estimate the unknown aerodynamic acceleration online. On this basis, maneuver estimation is used to modify the prediction step of extended Kalman filter to overcome the problem of filter divergence caused by model error. Finally, considering the time delay error of the maneuver observer, a suboptimal fading factor is introduced into the iterative process of the extended Kalman filter to improve the robustness of the filter tracking. Compared with typical tracking methods such as strong tracking filter, augmented state Kalman filter and interactive multiple model filter, the proposed method can achieve good stability and tracking accuracy with lower computing time without prior information of maneuvering target.

Key words: Hypersonic glide vehicle, Maneuvering target tracking, State estimate, Maneuver observer, Kalman filter