航空学报 > 2018, Vol. 39 Issue (10): 322247-322247   doi: 10.7527/S1000-6893.2018.22247

幅度及多普勒信息辅助的多目标跟踪算法

彭华甫1,2, 黄高明1, 田威1, 邱昊1   

  1. 1. 海军工程大学 电子工程学院, 武汉 430033;
    2. 中国人民解放军92773部队, 温州 325807
  • 收稿日期:2018-04-26 修回日期:2018-07-27 出版日期:2018-10-15 发布日期:2018-08-01
  • 通讯作者: 黄高明 E-mail:hgaom_paper@163.com
  • 基金资助:
    中国博士后科学基金(2017M613370)

Multi-target tracking algorithm based on amplitude and Doppler information

PENG Huafu1,2, HUANG Gaoming1, TIAN Wei1, QIU Hao1   

  1. 1. College of Electronic Engineering, Naval University of Engineering, Wuhan 430033, China;
    2. Unit 92773 of PLA, Wenzhou 325807, China
  • Received:2018-04-26 Revised:2018-07-27 Online:2018-10-15 Published:2018-08-01
  • Supported by:
    China Postdoctoral Science Foundation (2017M613370)

摘要: 针对现有的多目标跟踪(MTT)算法在强杂波环境下会出现严重的性能衰退,利用目标的幅度及多普勒特征,基于随机有限集(RFS)理论,提出了一种幅度及多普勒信息辅助的δ广义标签多伯努利(ADI-δ-GLMB)滤波器。首先,利用幅度及多普勒信息对目标状态进行扩展,在此基础上建立了新的量测似然函数。然后,基于δ-GLMB滤波器框架推导了新的更新方程。最后,进行仿真验证,结果表明:强杂波下,ADI-δ-GLMB滤波器跟踪性能明显优于δ-GLMB滤波器,估计精度及稳定性更高,且计算量更低;同信息辅助的概率假设密度(PHD)、多目标多伯努利(MeMBer)滤波器相比,ADI-δ-GLMB滤波器状态估计精度更高。

关键词: 多目标跟踪, 随机有限集, 幅度信息, 多普勒信息, 标签多伯努利

Abstract: To address the severe performance degradation of the existing Multi-Target Tracking (MTT) in high clutter environment, a novel Amplitude and Doppler Information assisted δ-Generalized Labeled Multi-Bernoulli (ADI-δ-GLMB) filter is proposed based on the theory of Random Finite Set (RFS). Firstly, the amplitude and Doppler information isintroduced to expand the target state, and a new function of measurement likelihood is established on this basis. Then, a new update equation is deduced based on the framework of the δ-GLMB filter. Finally, the simulation results show that the tracking performance of the proposed ADI-δ-GLMB filter is significantly better than that of the standard δ-GLMB filter, displaying higher estimation accuracy and stability, and a lower computational cost in dense clutter environment. In addition, the estimation accuracy of the proposed ADI-δ-GLMB filter is higher than that of the information assisted Probability Hypothesis Density (PHD) and Multi-target Multi-Bernoulli (MeMBer) filter.

Key words: multi-target tracking, random finite set, amplitude information, Doppler information, labeled multi-Bernoulli

中图分类号: