摘要
数据关联是多目标跟踪系统的核心 ,雷达目标的极化信息虽然不能直接用于目标航迹更新 ,但可以用来有效区分不同目标和杂波 ,进而减小量测与航迹关联的不确定性。在获取了目标的极化信息的前提下 ,根据极化 -几何参量综合距离提出广义“最近邻”数据关联算法。仿真结果表明 ,广义“最近邻”算法性能优于传统的“最近邻”
Data association plays an important role in multiple target tracking. Target's polarization information can't update target' track, but it can discriminate between different targets and clutter. Accordingly, the nondeterminacy of association is reduced. Given target's polarization information, a generalized ”Nearest Neighbor” algorithm is proposed basing on the extended distance defined in two dimensional domain of polarization and space. Simulations verified that the generalized ”Nearest Neighbor” algorithm outperforms that of the conventional ”Nearest Neighbor”.
出处
《现代雷达》
CSCD
北大核心
2004年第12期30-32,36,共4页
Modern Radar
基金
全国优秀博士论文资助 ( 0 810 0 10 1)
关键词
多目标跟踪
数据关联
极化
广义“最近邻”
multiple target tracking,data association,polarization,generalized “nearest neighbor”
