A New Adaptive Method for the Ballistic Trajectory Extrapolation
Jing Zhou · Journal of Information and Computational Science · 2014
Kalman fllter is one of the most commonly used algorithms in ballistic trajectory extrapolation system. The conventional Kalman fllter is adapted to discrete linear system, but it cannot deal with problems caused by inaccurate modeling error. Both linearization error and modeling error may cause instability and divergence. In this paper, we introduce a fading factor into the quasi-linear optimum smoothing Kalman flltering to keep the fllter process stable. Quasi-linear optimum smoothing fllter is used to reduce the linearization error. And the fading factor is then applied to the predicted covariance matrix. This approach is tested by a ballistic trajectory extrapolation system. Simulation results are presented as a demonstration of the efiectiveness of the proposed method. The results show a dramatical improvement to the fllter performance and the capability of restraining the flltering divergence.