结合聚类的GM-PHD滤波器辐射源群目标跟踪  

Emitter group targets tracking using GM-PHD filter combined with clustering

作　　者：朱友清[1] 周石琳[1] 高贵[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院,湖南长沙410073

出　　处：《系统工程与电子技术》2015年第9期1967-1973,共7页Systems Engineering and Electronics

摘　　要：群目标跟踪是一种情况更为复杂的多目标跟踪问题,由于军事辐射源目标经常出现雷达关机的情况,因此常用的多目标跟踪方法对于这类辐射源群目标的跟踪效果并不理想。为此,结合聚类技术提出了一种改进的高斯混合概率假设密度(Gaussian mixture-probability hypothesis density,GM-PHD)滤波器跟踪方法。该方法在GM-PHD滤波器的更新过程中,通过引入群中心产生的虚拟量测信息以提高目标跟踪性能,但不进行量测集划分。获得单一个体目标的估计状态后利用Jensen-Shannon divergence计算其相似度,然后再对估计目标进行聚类以实现群目标的跟踪。最后通过对相邻时刻的群中心轨迹点进行关联匹配,从而获得群目标的完整运动轨迹。仿真实验结果表明,所提方法能够对辐射源群目标进行有效跟踪,并具有较好的目标跟踪性能。Group targets tracking is a more complex problem of multi-target tracking. Because the military emitter targets often turn off the radar, the traditional tracking methods do not perform well for these emitter group targets. A modified Gaussian mixture-probability hypothesis density （GM-PHD） filter combined with clustering technology is proposed. In the update process of the GM-PHD filter, the proposed method introduces the dummy measurements generated by the group centers to improve the tracking performance, rather than par- titions the measurement set. After estimating the single target statements, the Jensen-Shannon divergence is used to compute their similarities. Then, the estimated targets are clustered to achieve the group tracking. Fi- nally, the track points of the group centers in adjacent time are connected to obtain the entire trajectories of the group targets. Experiment results show that the proposed method can effectively track the emitter group targets and performs better in the simulated scenarios.

关 键 词：群目标跟踪 高斯混合概率假设密度滤波器 聚类 航迹提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]