中文摘要：

多目标数据关联时,联合概率数据互联（JPDA）算法是最常用的方法之一,与最优的贝耶斯算法需要对当前时刻以前的所有确认量测集合进行研究相比,其只对最新的确认量测集合进行研究,因此是次优的贝叶斯算法。为进一步提高JPDA算法的性能,基于最优贝叶斯算法的理论,将包含目标历史信息的速度信息引入JP-DA算法过程中,增加了近距离平行运动目标的正确关联次数,并提高交叉运动目标关联精度。

英文摘要：

In multi-target tracking data association, joint probabilistic data association （JPDA） is one of the most popular algorithms. Compared with the optimal Bayes algorithm which considers all the measurements before current time, in the JPDA algorithm only the recent measurements are useful, so it is a sub-optimal Bayes algorithm. To improve the performance of JPDA, based on the theory of optimal Bayes algorithms, the velocity information which is based on history information is introduced, the new JPDA algorithm can increase the times of right association when close targets move in parallel, and enhance the precision of association when targets move crosswise.