一种改进的雷达弱目标检测前跟踪算法

An Improved Algorithm of Track-before-detection for Radar Weak Target

针对基于粒子滤波的检测前跟踪算法(PF-TBD)采样盲目性的问题,提出基于辅助粒子滤波的检测前跟踪算法(APF-TBD).该算法的重要性密度函数是一种依靠过去状态与系统最近观测量的密度函数,进一步使权值变化稳定,最终使得均方根误差较小,得到的粒子更逼近真实值,并将其应用于常规雷达对弱目标的检测与跟踪.仿真结果表明,该算法比PF-TBD有更好的检测效果与跟踪效果.

Aimed at the drawbacks ofresampling blindness in the algorithm of track-before-detection based on particle filters, the APF-TBD was proposed. The importance density function of the algorithm is a joint density function that depends on the past state and the most recent observations of the system, which results in the change of weight being robust and the mean square root error （RMSE） least, and then the acquired particle is even more approaching to the real value. And the algorithm can be applied to the detection and track of weak target in conventional radars. The simulation results show that the detection and track effect of the improved method is much better than that of the PF-TBD.