高斯厄米特粒子PHD被动测角多目标跟踪算法  

Gauss-Hermite particle PHD filter for bearings-only multi-target tracking

作　　者：杨金龙[1,2] 姬红兵[1] 刘进忙[3] 

机构地区：[1]西安电子科技大学电子工程学院,陕西西安710071 [2]江南大学物联网工程学院,江苏无锡214122 [3]空军工程大学导弹学院,陕西三原713800

出　　处：《系统工程与电子技术》2013年第3期457-462,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(60871074)资助课题

摘　　要：针对传统粒子概率假设密度(probability hypothesis density,PHD)滤波跟踪被动多目标时,估计精度不高,且存在粒子退化,容易导致滤波器发散的问题,提出一种新的被动多目标跟踪算法———高斯厄米特粒子PHD滤波算法。该算法采用一族高斯厄米特滤波产生的高斯分布拟合更优的重要性密度函数,充分考虑了当前时刻的最新量测,并将该方法融入高斯混合粒子PHD(Gaussian mixture particle PHD,GMP-PHD)滤波框架中,在解决观测非线性的同时,有效提高了被动多目标的跟踪精度。实验结果表明,该算法较传统的GMP-PHD滤波算法具有更高的状态估计精度,且有效降低了目标的失跟率。Taking into consideration the shortcomings of the traditional particle probability hypothesis density （PHD） filter algorithm for passive multi-target tracking, such as low accuracy, particle degradation, filter divergence, an improved multi-target tracking algorithm is proposed. In the proposed algorithm, the better importance density function is approximated by some new Gaussian distribution produced by a bunch of Gauss-Hermite filters, and the latest measurements are fully utilized. The Gauss-Hermite filters are integrated into the framework of Gaussian mixture particle PHD （GMP-PHD）, which solves the nonlinear problem and improves the accuracy of the proposed algorithm for passive multi-target tracking. Simulations show that the proposed algorithm has higher precision than the conventional GMP-PHD method, and it effectively decreases the loss rate of target estimates.

关 键 词：随机有限集 目标跟踪 概率假设密度 高斯厄米特滤波 

分 类 号：TN953[电子电信—信号与信息处理]