一种基于多传感器PHD滤波的非协同探测目标跟踪算法  

A Target Tracking Method for Non-cooperative Detection Systems Based on Multi-sensor PHD Filter

作　　者：许博 梁龙 欧阳成[1] 房汉林 XU Bo;LIANG Long;OUYANG Cheng;FANG Hanlin(Southwest China Research Institute of Electronic Equipment,Chengdu 610036,China;Military Representative Office of Central Military Commission Equipment Development Department,Chengdu 610036,China)

机构地区：[1]中国电子科技集团公司第二十九研究所,成都610036 [2]军委装备发展部驻某地区军事代表室,成都610036

出　　处：《电子信息对抗技术》2023年第4期51-57,共7页Electronic Information Warfare Technology

摘　　要：针对多传感器概率假设密度(Probability Hypothesis Density,PHD)滤波算法应用于多站非协同探测系统时,受杂波影响较大,且运算复杂度过高的问题,提出一种基于多传感器PHD滤波的非协同探测目标跟踪算法。首先,在多传感器PHD滤波的基础上,将非协同探测中特有的杂波对消剩余参数引入杂波密度的估计过程中,提高了算法的抗干扰能力。其次,将不同站点的空域覆盖引入多传感器PHD滤波的预测过程,从而不再需要考虑所有可能的排列组合所产生的高斯分量交叉项,大幅提高了运算效率。仿真试验对比了传统多传感器PHD滤波和改进算法的目标跟踪性能。结果表明,改进算法具有较强的环境适应能力,在多站非协同探测系统中具有良好的工程应用前景。Aiming at the problem that the multi-sensor probability hypothesis density(PHD)filtering algorithm is greatly affected by clutter and the computational complexity is too high when it is applied to the multi-sensor non-cooperative detection systems,a non-cooperative detection target tracking algorithm based on multi-sensor PHD filtering is proposed.Firstly,on the basis of multi-sensor PHD filtering,the residual parameters of clutter cancellation unique to non-cooperative detection are introduced into the estimation process of clutter density,which improves the anti-interference ability of the algorithm.Secondly,the spatial coverage of different stations is introduced into the prediction process of the multi-sensor PHD filter,so that it is no longer necessary to consider the Gaussian components generated by all possible combinations,which can greatly improve the computational efficiency.The target tracking performance of traditional multi-sensor PHD filter and improved algorithm is compared by simulation.The results show that the improved algorithm has strong adaptability to the environment and has a good application prospect in the multi-static non-cooperative detection systems.

关 键 词：随机集 多传感器 概率假设密度滤波 非协同探测 

分 类 号：TN953[电子电信—信号与信息处理]