干扰环境下数目可变的多目标跟踪

摘要

针对环境中干扰因素对目标跟踪性能的影响,提出了一种跟踪数目可变的多目标方法.该方法通过聚类算法来判断目标的出现或消亡,并获得目标数目,然后利用基于无迹卡尔曼的廉价联合概率数据关联算法获得量测与已知目标轨迹的互联概率,并实现状态传递.与传统的廉价联合概率数据关联算法相比,该算法在干扰环境下跟踪性能更优,可处理目标数目可变的多目标跟踪问题.%In this paper, a passive bistatic radar system using television as opportunistic illuminator isrnpresented. When considering the existence of clutters, a new method is proposed to track a variable numberrnof targets in a cluttered environment. A clustering algorithm is first introduced to judge whether a new targetrnappears or an existing target vanishes, and then the UKF based cheap joint probabilistic data associationrnmethod (UKFCJPDA) is used to track the targets. Numerical simulation shows good performance of thisrnalgorithm.