基于K-中心点聚类的模糊航迹关联算法

Fuzzy track association algorithm based on K-center clustering

为提高目标航迹相交和近距平行状态时航迹关联的正确率,提出了一种基于K-中心点聚类的模糊航迹关联算法。该算法基于K-中心点聚类算法,将系统航迹作为聚类中心,采用局部航迹与系统航迹关联的策略,为描述航迹间的相似性,采用模糊分析方法,综合考虑各个因素的影响,构造模糊关联矩阵,并利用历史信息和先验知识进行航迹关联。仿真表明该算法在航迹相交状态下,相交时刻关联正确率比K-medoids聚类算法提高5%左右,近距平行状态下关联正确率的收敛速度优于K-medoids聚类算法。

To improve the probability of correct track association when tracks of targets intersect or are in parallel within near range, a fuzzy track association algorithm based on K-center clustering was presented. With system track as clustering center, the algorithm associated sensor tracks with system tracks, and thus improved the efficiency of the algorithm considerably. Fuzzy set theory is applied in the algorithm to describe the similarity of tracks and evaluate the influence of various factors. Accordingly a fuzzy association matrix was built and history information and priori knowledge were used to associate tracks. The simulation results demonstrate five percent higher accuracy at the intersection point for cross tracks and higher convergence rate of the accurate rate of association in the case of near-range parallel tracks than K-merdoids clustering algorithm.