J4 ›› 2015, Vol. 42 ›› Issue (5): 98-104.doi: 10.3969/j.issn.1001-2400.2015.05.017

• 研究论文 • 上一篇    下一篇

未知杂波环境的GM-PHD平滑滤波器

李翠芸1;江舟2;李斌2;周旋3   

  1. (1. 西安电子科技大学 电子工程学院,陕西 西安  710071;
    2. 中国人民解放军95972部队,甘肃 酒泉  735018;
    3. 中国人民解放军96217部队,海南 三亚  572011)
  • 收稿日期:2014-05-16 出版日期:2015-10-20 发布日期:2015-12-03
  • 通讯作者: 李翠芸
  • 作者简介:李翠芸(1976-),女,副教授,博士,E-mail: cyli@xidian.edu.cn.
  • 基金资助:

    国家自然科学基金资助项目(61372003);国家自然科学基金青年基金资助项目(61101246,61301289);中央高校基本科研业务费专项资金资助项目(K5051202014)

Gaussian mixture PHD smoothing filter in unknown clutter

LI Cuiyun1;JIANG Zhou2;LI Bin2;ZHOU Xuan3   

  1. (1. School of Electronic Engineering, Xidian Univ., Xi'an  710071, China;
    2. The Chinese PLA 95972 Troops, Jiuquan  735018, China;
    3. The Chinese PLA 96217 Troops, Sanya  572011, China)
  • Received:2014-05-16 Online:2015-10-20 Published:2015-12-03
  • Contact: LI Cuiyun

摘要:

针对未知杂波环境下的多目标跟踪问题,提出一种未知杂波环境下的高斯混合概率假设密度前向后向平滑算法.该算法首先利用有限混合模型对杂波强度进行估计,克服了多目标跟踪中概率假设密度滤波器在杂波与先验知识不匹配情况下滤波性能急剧下降的缺点; 其次采用平滑递归,利用多个量测数据对滤波值进行平滑,进而减小目标的位置误差.仿真结果表明,这种算法在未知杂波环境下具有较好的跟踪性能,且优于未进行平滑的未知杂波高斯混合概率假设密度滤波器.

关键词: 未知杂波, 高斯混合概率假设密度, 平滑, 多目标跟踪

Abstract:

Aiming at the Multi-target tracking in the unknown clutter environment, this paper proposes a Gaussian Mixture Probability Hypothesis Density (GM-PHD) forward-backward smoothing algorithm, which improves the poor performance of the PHD filter when the clutter model and the prior knowledge are mismatching by estimating the clutter intensity with the finite mixture model. The forward-backward smoothing recursions are applied to improve the state estimation accuracy. Simulation results show that the proposed algorithm performs well in the unknown clutter environment and better than the conventional Gaussian Mixture PHD Filter without smoothing processing in the unknown clutter environment.

Key words: unknown clutter, Gaussian mixture probability hypothesis density, smoothing, multitarget tracking

Baidu
map