无线电工程

目标航迹聚类能够从海量的航迹数据中发现目标的潜在运动轨迹,是态势数据挖掘、行为意图分析的重要技术手段。提出一种能够对目标航迹数据进行聚类分析、发现目标潜在运动轨迹的算法,以航迹线段作为分析对象,根据垂直距离、平行距离和角度距离综合判断2条航迹线段之间的距离,将距离相近的线段聚为一类。实验结果表明,该算法能够从大量航迹数据中发现目标运动的轨迹特征。

Clustering track data of moving objects is an important means for the analysis of situation data and speculation on objects' behavioral intentions. An algorithm of mining moving path data of objects for movement track patterns is presented in this paper. This algorithm is special in that it focuses on path segments,instead of the whole path,to implement mining patterns. Experimental results show that this approach is effective and scalable.