关键词: 自动控制技术, 多模型自适应估计, 不敏卡尔曼滤波, 数值鲁棒性, 脱靶量

Abstract:

A fast and efficient algorithm is proposed for multimodel adaptive estimation in the terminal interception of randomly maneuvering target scenario. The traditional multimodel adaptive estimation is simplified by means of aggregation and pruning, based on the exploitation of the special structure of the hypothesis space corresponding to a bank of elemental filters. To efficiently handle the nonlinearity, unscented Kalman filter is introduced into the design of elemental filter that is hypothesis dependent. Numerically robust implementation of the algorithm is presented to meet the need of practical application. Simulation results demonstrate the feasibility of the proposed algorithm.

Key words: automatic control technology, multiple model adaptive estimation(MMAE), unscented Kalman filter (UKF), numerical robustness, miss distance