基于二值粒子群优化的多目标协同检测与跟踪方法  

Multi- target Collaborative Detection and Tracking Based on Binary Particle Swarm Optimization

作　　者：刘钦[1] 

机构地区：[1]中国电子科技集团公司第二十研究所,西安710068

出　　处：《火控雷达技术》2015年第1期1-6,共6页Fire Control Radar Technology

基　　金：中央高校基本科研业务费专项基金资助项目(K5051202036)

摘　　要：针对如何根据目标被探测状态(被检测或者被跟踪)对有限的雷达资源进行分配的问题,本文将其转化为组合优化问题,提出了一种新颖的基于后验克拉美罗下界(PCRLB)-二值粒子群优化(BPSO)的雷达-目标自动分配算法。该算法采用PCRLB作为已跟踪目标的跟踪精度衡量标准,并将其与新生目标的检测概率构成BPSO的适应度函数,在最大化新生目标检测概率的条件下,最小化已跟踪的多个目标的PCRLB,自适应地为目标分配雷达完成恰当的探测(检测与跟踪)行为。仿真结果表明,该算法不仅能够及时检测新生目标,而且能够持续且优化跟踪已有目标,使网络的整体精度得到明显提高。In order to achieve effective assignment of limited radar source based on target status （detected or tracked） , a novel radar - target automatic assignment algorithm based on posterior Cramer - Rao lower bound （PCRLB） and binary particle swarm optimization （BPSO） is proposed. By applying PCRLB as tracking accuracy measurement standard of tracked target, and by combination of new target detection probability to obtain adaptation function of BPSO ; through maximization of the new target detection probability and minimization of PCRLB of multi- ple tracked targets, so as to assign multiple targets to radar adaptively for proper detection （ detect or track）. Simu- lation results verify that, by using this algorithm, new detected target can be found, and tracked target can be tracked continuously, therefore global accuracy of the network get improved significantly.

关 键 词：协同检测 协同跟踪 二值粒子群优化 

分 类 号：TN957.51[电子电信—信号与信息处理]