基于误差校准的多雷达粒子滤波检测跟踪算法  

A Multi-Radar Particle Filter Based on Error Registration for TBD Algorithm

作　　者：程小会[1] 陈霄[1] 薛安克[1] 

机构地区：[1]杭州电子科技大学通信信息传输与融合技术国防重点学科实验室,杭州310018

出　　处：《火力与指挥控制》2016年第10期6-9,共4页Fire Control & Command Control

基　　金：国家自然科学基金(61004119;61375011);国家"973"计划基金资助项目(2012CB821204)

摘　　要：基于粒子滤波的检测前跟踪方法是一种处理弱目标检测与跟踪的有效方法。使用多部雷达对同一目标进行观测,可以提高目标的检测概率和跟踪精度。但雷达系统误差不同,得到的目标量测信息不能直接进行融合。针对多个具有不同系统误差的雷达联合检测跟踪问题,通过将不同雷达的量测信息向前追溯,对误差进行校准,从而消除不同雷达对于同一粒子的量测误差,然后将粒子权重进行融合。仿真结果表明,在雷达具有不同系统误差的情况下,采用本算法可以有效提高目标的跟踪精度。Particle filter based track-before-detect algorithm is an efficient approach to detect andtrack weak target, the way that to observe the same target using multi-radars can improve the trackingaccuracy and the probability of target detection. But because different radars have different systemerrors, so the target information can’ t be directly integrated. To solve the problem of multiple withdifferent system errors of radar detection tracking problem, through the different radar measurementinformation going back processing of the error of the calibration, thereby eliminating different radarsfor particles with the same amount of measurement error, then the particle weight fusion. Simulationresults show that this algorithm can effectively improve the tracking accuracy of the target in the case ofradar with different system errors.

关 键 词：检测前跟踪 粒子滤波 同步雷达 融合检测 

分 类 号：TN953[电子电信—信号与信息处理]