基于信息融合的多目标单站无源跟踪新技术研究

跟踪起始与数据关联是多目标无源单站跟踪的关键技术.本文提出了一种基于目标多特征信息融合的自适应跟踪起始算法,通过构造多维动态可变的跟踪门,进行自适应跟踪起始检测,然后根据序列概率比检验准则进行轨迹确认.同时提出了一种基于多目标多特征信息融合的数据关联算法,首先通过定义多个特征数据关联度,将单个有效观测的多特征信息进行融合,再对多目标进行综合数据关联.计算机仿真表明,该跟踪起始算法能够快速有效地启动航迹,数据关联算法的性能要优于传统的NN方法和扩展的NN方法.

Study of New Techniques for Multiple Tartest Monostatic Passive Tracking Based on Information Fusion

Target initiation and data association are key techniques for multiple target monostatic tracking. This paper presents an adaptive tracking initiation algorithm based on multiple target information fusion. It constructs multiple dimension dynamic variable range gates for adaptive tracking initiation detection, then makes track verification based on sequence probability rate detection criteria. Moreover, this paper gives a data association algorithm based on multiple target multiple signature information fusion algorithm. It first defines multiple signature data association degree to fuse single effectively measured multiple signature information, and then makes compounded data association of multiple targets. Computer simulations show that the tracking initiation algorithm can initiate fast and effectively the track and the data association algorithm outperforms the traditional NN method and the expanded NN method.