一种适于工程应用的近似MSJPDA算法

为提高复杂数据融合系统中的航迹关联正确率，在ZHOU B的DC和AC算法基础上提出一种新的近似多传感器多目标联合概率数据关联算法，它以一个目标为中心的近似聚为构造关联事件的起点，并在计算中将DC和AC结合得到的一种全部的点迹－航迹关联算法，在杂波下目标密集，航迹复杂的数据融合系统中进行实验，对关联正确率，并联时耗等与最近邻法进行了比较，效果较好，它能为有效提高目标点迹－航迹的关联正确率，在计算时耗上较完全联合概率法少得多，能满足工程中实时性的要求。

Approximate Multi-Sensor Multi-Target Joint Probabilistic Data Association Algorithm Applicable to Engineering

To improve the correct association rate of track in complex data fusion system, a new association algorithm-approximate multi-sensor multi-target joint probabilistic data association (AMSJPDA) is presented in the paper based on DC and AC brought by ZHOU B. AMSJPDA is a neighbor plot-track association algorithm in which the approximate set of a target is made as the start-point to construct associations and DC is united with AC. It is tested in the data fusion system of compacted targets under scatter wave and complex track. Compared to nearest neighbor, it shows better effects of the correct association rate and association time. It can improve the correct association rate and target plot-track and demand less time needed than joint probabilistic data association. The AMSJPDA can meet the requirements of real-time in engineering.