Abstract: The hypersonic target in near space has the characteristics of high maneuverability, high speed and high nonlinearity,which usually result in some problems in the process of target tracking, such as the reduced tracking accuracy and the filter diverging. To solve the problems, an interactive multiple model cubature particle filter algorithm is proposed. This algorithm uses the interactive multiple model algorithm to track the hypersonic target in near space, and uses the cubature particle filter algorithm to predict the state. The simulation results show that, compared with the interactive multiple model Kalman filter algorithm and interactive multiple model particle filter algorithm, this algorithm has better tracking performance for the hypersonic target in near space.

Key words: near space hypersonic target, high maneuverability, interactive multiple model, cubature particle filter, target tracking, nonlinearity, X-51 aircraft

摘要： 临近空间高超声速目标具有机动特性强、轨迹变化快、强非线性等特点,在目标跟踪的过程中,易出现跟踪精度降低、滤波发散的问题。针对该问题,提出了一种交互式多模型容积粒子滤波算法。使用交互式多模型算法来对临近空间高超声速目标进行跟踪,使用容积粒子滤波算法对目标进行滤波预测。仿真结果表明,该算法跟踪性能优于交互式多模型卡尔曼滤波算法和交互式多模型粒子滤波算法,对临近空间高超声速目标有更好的跟踪效果。

关键词: 临近空间高超声速目标, 机动特性强, 交互式多模型, 容积粒子滤波, 目标跟踪, 非线性, X-51A飞行器