摘要
本文提出了一种基于局部估计融合的多传感器多目标（ＭＳＭＴ）数据关联算法，它包括粗关联和精关联两个过程．本文首先将单传感器中基于２－３树搜索的关联算法［１］推广到ＭＳＭＴ情况，在粗关联基础上，给出了主传感器位置估计与其它传感器位置估计之间的关系，通过最近邻联合概率数据关联算法（ＮＮＪＰＤＡ）完成了二者局部估计之间的精关联．最终将所有来自同一目标的局部估计相融合获得全局估计．与已有的分布ＪＰＤＡ算法相比，本文算法在性能基本接近的情况下有效地减小了运算量．
A fast algorithm for MSMT data association, which includes initial association and fine association, is presented based on the fusion of local estimations. The data association algorithm[1] based on ternary searching is extended to MSMT case. On the basis of it, the relation between local position estimations at the main sensor and ones at others is given. The fine association is completed by using nearst neighbor JPDA(NNJPI)A). All local estirnations from the same targets are fused to obtain the global estiniations. A comparison of the algorithm here with the existing disiributed JPDA one is made, which shows that the algorithm here effeciently reduces computational cost with basic approximate performance.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1996年第6期62-66,共5页
Acta Electronica Sinica
基金
国防预研基金
关键词
多传感器
多目标跟踪
数据处理
数据融合
Multisensor, Multitarget tracking, Data association, Data fusion
