基于数据关联的多雷达点迹融合算法研究  

Research on Multi-radar Point Track Fusion Algorithm Based on Data Association

作　　者：张昕 张博文 张玉萍 李军侠 燕瑞超 Zhang Xin;Zhang Bowen;Zhang Yuping;Li Junxia;Yan Ruichao(System Engineering Research Institute of CSSC,Beijing 100094,China)

机构地区：[1]中国船舶工业系统工程研究院

出　　处：《计算机测量与控制》2019年第8期207-211,216,共6页Computer Measurement &Control

摘　　要：在典型应用场景下,针对传统数据关联算法易于产生目标跟踪丢失的问题,提出了并行多传感器联合概率数据关联算法(PMSJPDA),给出行多传感器联合概率数据关联算法的实现流程,并对该算法进行理论分析;最后采用两部典型雷达构成PMSJPDA数据融合系统,对其算法的目标跟踪精度进行1 000次蒙特卡洛仿真,仿真统计结果表明,在目标跟踪全程阶段,融合后的目标跟踪精度比单部雷达跟踪精度提高约6%,在目标跟踪稳定阶段,融合后的目标跟踪精度比单部雷达跟踪精度提高约10%,即所提出的PMSJPDA算法可有效提高对目标的跟踪性能,尤其是目标跟踪进入稳定阶段。In the typical application scenario,aiming at the problem that the traditional data association algorithm is easy to generate target tracking loss,the parallel multi-sensor joint probability data association(PMSJPDA)realization process is proposed,and the algorithm is theoretically studied.Finally,two typical radars are used to construct a data fusion system;the performance of the data fusion algorithm is simulated and verified in the shipboard application scenarios.Simulation results show that tracking whole segment the target tracking accuracy of the fusion is 6%higher than the single radar tracking.Simulation results show that tracking stable segment the target tracking accuracy of the fusion is 10%higher than the single radar tracking.The proposed PMSJPDA algorithm can effectively improve the tracking performance.

关 键 词：点迹融合 联合概率数据关联 组网雷达 融合算法 并行结构 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]