基于航迹概率假设密度的多传感器多目标跟踪  

Multi-sensor multi-target tracking with trajectory probability hypothesis density

作　　者：王志伟 刘永祥 杨威 卢哲俊 WANG Zhiwei;LIU Yongxiang;YANG Wei;LU Zhejun(College of Electronic Science and Technology,National University of Defense Technology,Changsha 410073,China)

机构地区：[1]国防科技大学电子科学学院,湖南长沙410073

出　　处：《系统工程与电子技术》2024年第2期526-533,共8页Systems Engineering and Electronics

基　　金：国家自然科学基金(61901498,61871384,61921001);湖南省优秀博士后创新人才基金(2020RC2043)资助课题。

摘　　要：针对基于概率假设密度(probability hypothesis density,PHD)的分布式多传感器多目标跟踪(distributed multi-sensor multi-target tracking,DMMT)存在无法形成航迹、计算复杂度高、目标漏检等问题。本文基于航迹PHD后验估计提出了一种DMMT方法。为此,首先构建了各节点估计航迹间相似性度量矩阵,并采用匈牙利算法实现最优航迹匹配;其次采用协方差逆准则对关联航迹实现并行融合;最后基于概率生成泛函推导了一种鲁棒的DMMT方法。仿真实验验证了所提算法在目标状态估计精度、计算有效性和实时性方面的优势。Aiming at the problems that the distributed multi-sensor multi-target tracking(DMMT)method based on probability hypothesis density(PHD)cannot form a track,the computational complexity is high,and miss-detections.In this paper,a DMMT method is developed with trajectory PHD posterior estimation.Firstly,the similarity measure matrix between the estimated trajectories of each node is constructed,and the optimal trajectory matching is achieved by using the Hungarian algorithm.Secondly,the covariance intersection rule is used to achieve parallel fusion for the associated trajectories.Finally,a robust DMMT method is derived based on probabilistic generative functionals.Simulation experiments verify the advantages of the proposed algorithm in terms of multi-target state estimation accuracy,computational efficiency and real-timeliness.

关 键 词：航迹概率假设密度 最优航迹匹配 广义协方差逆 概率生成泛函 

分 类 号：TN957[电子电信—信号与信息处理]