电子学报 ›› 2014, Vol. 42 ›› Issue (10): 1887-1893.DOI: 10.3969/j.issn.0372-2112.2014.10.003

• 学术论文 • 上一篇    下一篇

角度传感器网络多目标定位的数据关联算法

李猛1, 王智1, 李元实1, 鲍明2   

  1. 1. 浙江大学工业控制技术国家重点实验室, 浙江杭州 310027;
    2. 中国科学院声学研究所, 北京 100190
  • 收稿日期:2013-07-15 修回日期:2013-12-06 出版日期:2014-10-25
    • 通讯作者:
    • 王智
    • 作者简介:
    • 李 猛 男,1989年2月生于吉林长春.2012年毕业于上海交通大学自动化系,现为浙江大学控制科学与工程学系硕士研究生.主要研究方向为传感器网络定位追踪. E-mail:limeng198921@mail.com
    • 基金资助:
    • 国家自然科学基金 (No.NSFC61273079); 中国科学院战略性先导科技专项 (No.XDA06020300,No.XDA06020201); 工业控制技术国家重点实验室开放课题 (No.ICT1206,No.ICT1207)

Data Association in Multi-Target Localization Using Bearing-Only Sensor Networks

LI Meng1, WANG Zhi1, LI Yuan-shi1, BAO Ming2   

  1. 1. State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou, Zhejiang 310027, China;
    2. Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China
  • Received:2013-07-15 Revised:2013-12-06 Online:2014-10-25 Published:2014-10-25
    • Supported by:
    • National Natural Science Foundation of China (No.NSFC61273079); Chinese Academy of Sciences Strategic Pilot Project (No.XDA06020300, No.XDA06020201); Open Project of State Key Laboratory of Industrial Control Technology (No.ICT1206, No.ICT1207)

摘要:

在基于方向角等被动式的多目标监测中,由于面临量测—目标匹配关系未知,现有定位算法很难实时给出目标的精确位置,尤其是在方向角量测误差存在的情况下.针对不可靠角度量测传感器网络下的多目标定位问题,分析和设计了无目标先验信息下的量测关联算法.通过分析量测误差对关联算法的影响,给出所得多目标定位来源于真实目标概率的理论推导以及相关门限选取方法,并应用在此基础上设计算法,给出最优的多目标位置组合.该关联算法在引入各传感器量测信息的同时更新门限,以此保证多重定位是真实目标定位的可靠性.仿真结果表明,所提出的数据关联算法于所示情形下均具有较好的性能,在多目标定位中能捕获大部分目标,且计算量较低.

关键词: 数据关联, 基于方向角的传感器网络, 多目标定位, 感知概率

Abstract:

Multi-target bearing-only localization on sensor network suffers from unknown association between measurements and targets, making it difficult to achieve real-time localization, especially with the existence of bearing measurement errors.A data association method is designed for multi-target localization to deal with the unknown target initial states and complex measuring phenomena.Theoretical deduction of sensing probability and gate threshold selection is given on the basis of measurement error analysis.A data association method based on scalable gate threshold is proposed to estimate target when the measurement is importing to the data fusion center, ensuring the probability that intersection of bearing measurement is real target localization.Simulation results show that the proposed method works well under different settings and the required acceptable computation load.

Key words: data association, bearing-only sensor networks, multiple-target localization, sensing probability

中图分类号: