航空学报 > 2023, Vol. 44 Issue (S2): 729708-729708   doi: 10.7527/S1000-6893.2023.29708

基于图模型的无人集群同步自定位与相对定位

熊骏1,2(), 解相朋1, 熊智3, 庄园2, 郑宇4   

  1. 1.南京邮电大学 物联网学院,南京 210042
    2.武汉大学 测绘遥感信息工程国家重点实验室,武汉 430079
    3.南京航空航天大学 自动化学院,南京 211106
    4.中国兵器工业导航与控制技术研究所,北京 100089
  • 收稿日期:2023-10-12 修回日期:2023-11-14 接受日期:2023-12-01 出版日期:2023-12-08 发布日期:2023-12-07
  • 通讯作者: 熊骏 E-mail:xiongjun@njupt.edu.cn
  • 基金资助:
    国家自然科学基金(62203228);南京邮电大学引进人才自然科学研究启动基金(NY221137);航空科学基金(ASFC-2022Z0220X9001);武汉大学测绘遥感信息工程国家重点实验室开放研究基金(22P01);中国博士后科学基金(2023M742216)

Synchronized self⁃localization and relative⁃localization of unmanned swarms based on graph model

Jun XIONG1,2(), Xiangpeng XIE1, Zhi XIONG3, Yuan ZHUANG2, Yu ZHENG4   

  1. 1.College of Internet?of?Things,Nanjing University of Posts and Telecommunications,Nanjing 210042,China
    2.State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China
    3.College of Automation,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China
    4.Navigation and Control Technology Research Institute of China North Industries Group Corporation,Beijing 100089,China
  • Received:2023-10-12 Revised:2023-11-14 Accepted:2023-12-01 Online:2023-12-08 Published:2023-12-07
  • Contact: Jun XIONG E-mail:xiongjun@njupt.edu.cn
  • Supported by:
    National Natural Science Foundation of China(62203228);Natural Science Research Start-up Foundation of Recruiting Talents of Nanjing University of Posts and Telecommunications(NY221137);Aeronautical Science Foundation of China(ASFC-2022Z0220X9001);Open Research Fund Program of LIESMARS, Wuhan University (22P01);China Postdoctoral Science Foundation(2023M742216)

摘要:

协同定位是一种在导航基础设施受限环境下提高无人集群定位精度的技术,包含无人载体的自定位和相对定位2方面。其中,自定位用于获取自身绝对位姿,相对定位用于获取载体间的相对位姿。然而,现有协同定位方法通常解决其中一方面问题而忽视另一方面,这不仅限制了协同定位的工程应用价值,也忽视了集群运动与个体运动间约束关系,从而降低协同估计性能。为解决该问题,本文提出了一种可同时进行自定位和相对定位的协同估计方法。以概率约束关系构建概率图模型,用于描述自定位状态、相对定位状态和集群量测间的概率关系;利用高斯信息形式在图模型中进行定位状态边缘概率求解,实现图模型高效计算。实验结果表明,本方法不仅能够实现自定位和相对定位的同步估计,且估计精度优于传统协同定位方法,自定位误差与相对定位误差分别降低59.2%和30.3%,为集群协同导航定位提供了一种新的融合框架。

关键词: 无人集群, 协同定位, 图模型, 相对定位, 自定位

Abstract:

Cooperative localization, including self-localization and relative-localization of unmanned carriers, is a technology that improves the localization accuracy of unmanned swarms in navigation infrastructure-limited environments. Self-localization is used to obtain the absolute pose of oneself, and relative-localization is used to obtain the relative pose between carriers. However, existing cooperative localization methods usually solve one aspect of the problem while ignoring the other, which not only limits the engineering application value of cooperative localization but also ignores the constraint relationship between swarm motion and individual motion; thereby, reducing cooperative estimation performance. To solve this problem, this paper proposes a Simultaneous Self and Relative Localization (SSRL) method that can parallelly perform self-localization and relative-localization of collaborative swarm. A probability graph model is constructed with probability constraint relationships to describe the probability relationship between self-localization states, relative localization states, and swarm measurements. Moreover, Gaussian information form is used to estimate the marginal probability of localization states in the graph model via an efficient approach. The experimental results show that SSRL can not only achieve simultaneous estimation of self-localization and relative-localization, but also has better estimation accuracy than traditional cooperative localization methods. The self-localization error and relative-localization error are reduced by 59.2% and 30.3%, respectively, indicating that SSRL provides a new fusion framework for swarm cooperative localization systems.

Key words: unmanned swarm, cooperative localization, graph model, relative-localization, self-localization

中图分类号: