基于随机集的传感器网络多目标航迹跟踪研究  

Multi-target Trajectories Tracking in Sensor Networks with Random Finite Set

作　　者：周红波[1] 万福[1] 李炳伟[1] 

机构地区：[1]海军指挥学院,南京211800

出　　处：《火力与指挥控制》2014年第6期17-21,25,共6页Fire Control & Command Control

基　　金：国家"八六三"计划基金资助项目(2007AA01Z309)

摘　　要：针对利用有限集统计理论对声学WSN下的多目标跟踪时,无法实现前后两个时刻目标状态的关联问题,利用最近邻方法和粒子标签法对WSN多目标的航迹跟踪问题进行了研究。仿真结果表明,两种算法都能够有效实现对多个目标航迹的跟踪;另外,粒子标签方法进行多目标跟踪时能够同时实现对多目标状态的估计和对多目标航迹的跟踪。To solve the problem that states between two successive times can＇t be associated when tracking multi-target in WSN using finite set statistic（FISST） theory,the multi-target trajectories tracking using nearest neighbor method and particle labeling method are studied. Simulations results show that two methods both can track multi-target trajectories effectively. In addition,the particle labeling method can estimate multi-target states and track multi-target trajectories at the same time when it is used for tracking multi-target.

关 键 词：无限传感器网络 多目标航迹跟踪 有限集统计理论 粒子滤波 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]