An Adaptive Kalman Filter for Near Space Hypersonic Vehicle Tracking

Abstract

Considering the characteristic of periodic ski-jump flight in the cruising stage of near space hypersonic vehicles, Sine-Jerk model is studied in this paper. Compared with other models, the Sine-Jerk model can further improve the matching degree of strongly maneuvering targets with periodicity. In practice, target maneuver angular rate is often unknown or even variable. When the actual maneuvering angular velocity deviates greatly from the preset value, the tracks filtering results will have deviation or even divergence. Therefore, in order to solve the above problems, this paper proposes a parameter adaptive Kalman filter algorithm, which can still have high tracking accuracy when the target angular rate changes. According to the filtering innovation, the algorithm judges whether the actual angular rate of the target matches the preset model and adaptively adjusts the filtering parameters by using the double-window detection method, so that the filtering gain matrix and the prediction error covariance can change adaptively with the change of the measurement residual. Compared with the Kalman filter algorithm with fixed tracking parameters, it is proved that the proposed algorithm has better filtering effect on the ski-jump maneuvering target in near space.