J4 ›› 2012, Vol. 39 ›› Issue (6): 49-54+135.doi: 10.3969/j.issn.1001-2400.2012.06.008

• 研究论文 • 上一篇    下一篇

一种多机动目标协同跟踪的博弈论算法

刘钦;刘峥;刘俊   

  1. (西安电子科技大学 雷达信号处理国家重点实验室,陕西 西安  710071)
  • 收稿日期:2011-08-02 出版日期:2012-12-20 发布日期:2013-01-17
  • 通讯作者: 刘钦
  • 作者简介:刘钦(1984-),女,西安电子科技大学博士研究生,E-mail: qinl@mail.xidian.edu.cn.
  • 基金资助:

    长江学者和创新团队发展计划资助项目(IRT0645);国家重大专项工程基础理论研究资助项目(G52809220262); 中央高校基本科研业务费专项资金资助项目(K5051202036)

Collaborative tracking algorithm for multiple maneuvering targets based on the game theory

LIU Qin;LIU Zheng;LIU Jun   

  1. (National Key Lab. of Radar Signal Processing, Xidian Univ., Xi'an  710071, China)
  • Received:2011-08-02 Online:2012-12-20 Published:2013-01-17
  • Contact: LIU Qin

摘要:

针对传感器网络中的动态跟踪问题,提出一种基于博弈论的多机动目标协同跟踪算法.首先利用交互多模型扩展粒子滤波估计网络中每个机动目标的状态;然后以滤波过程中获得的目标信息增益为衡量标准,促使跟踪精度未达系统要求的目标的代理发起谈判,在保证谈判双方利益最大化的前提下,通过博弈为谈判发起者争取更多的传感器对其所代表的目标进行跟踪.仿真结果表明,在非线性非高斯环境下,该方法与传统方法相比能够有效提高跟踪精度,动态分配传感器资源以实现协同跟踪.

关键词: 博弈论, 扩展卡尔曼粒子滤波, 交互多模型, 协同跟踪, 传感器网络

Abstract:

For the dynamic tracking problem in sensor networks, a new collaborative tracking algorithm for multiple maneuvering targets based on game theory is proposed. First, the interacting multiple model extended particle filtering algorithm is applied to obtain the estimated state by each sensor in the networks. Then, the target information gain is utilized as a basis for sensor negotiation. When the desired covariance level is updated during target tracking, a new negotiation will start to reallocate resources. The negotiation will allocate more sensors to the negotiation caller for its tracking under the condition of protecting the interests of both sides. Simulation results show that the proposed method can load to the desired tracking accuracy compared with traditional methods in a nonlinear non-Gaussian system.

Key words: game theory, extended Kalman particle filtering, interacting multiple model, collaborative tracking, sensor

中图分类号: 

  • TP391