一种新的自适应机动目标跟踪算法

在＂当前＂统计（CS）模型基础上,提出了一种新的机动目标自适应滤波算法,当前统计模型-修正强跟踪滤波（CS-MSTF）算法。新算法在保留＂当前＂统计模型及强跟踪滤波器（STF）对一般机动目标跟踪精度高的优点的同时,作出以下改进：针对强跟踪滤波器在机动部分获得完美性能的同时,非机动部分的精度却不理想的缺陷,对预测误差协方差及渐消因子的计算作出修正,同时改进机动部分和非机动部分的精度;将目前常用的估计误差协方差的计算公式采用更加可靠的Joseph公式,增强了数值的稳定性和算法的鲁棒性。蒙特卡罗仿真表明,新算法的性能优于当前统计模型-强跟踪滤波（CS-STF）算法,能够进行有效估计。

Abstract：

Based on the ＂current＂ statistical model,a new adaptive maneuvering target tracking algorithm,CS-MSTF,was proposed. The new algorithm,keeping the merits of high tracking precision that the ＂current ＂ statistical model and strong tracking filter（STF） have in tracking maneuvering target has made the modifications as such：First,STF has the defect that it achieves the perfert performance in maneuvering segment at a cost of the precision in non-naneuvering segment,so the new algorithm modifies the prediction error covariance matrix and the fading factor to improve the tracking precision both of the maneuvering segment and non-maneuvering segment; The estimation error covariance matrix was calculated using the Joseph form,which is more stable and robust in numerical. The Monte-Carlo simulation shows that the CS-MSTF algorithm has a more excellent performance than CS-STF and can esitmate efficiently.