通信学报 ›› 2023, Vol. 44 ›› Issue (6): 211-222.doi: 10.11959/j.issn.1000-436x.2023104

• 学术通信 • 上一篇    下一篇

基于生成模型的地磁室内高精度定位算法研究

马帅1,2,3, 裴科3, 祁华艳3, 李航4, 曹雯5, 王洪梅3, 熊海良6, 李世银3   

  1. 1 深地科学与工程云龙湖实验室,江苏 徐州 221116
    2 鹏城实验室,广东 深圳 518172
    3 中国矿业大学信息与控制工程学院,江苏 徐州 221116
    4 深圳市大数据研究院,广东 深圳 518172
    5 长安大学电子与控制工程学院,陕西 西安 710064
    6 山东大学信息科学与工程学院,山东 济南 250100
  • 修回日期:2023-04-12 出版日期:2023-06-01 发布日期:2023-06-01
  • 作者简介:马帅(1986- ),男,山东日照人,博士,鹏城实验室副研究员,主要研究方向为语义通信和通信定位一体化等
    裴科(1998- ),男,江苏宿迁人,中国矿业大学硕士生,主要研究方向为室内地磁定位
    祁华艳(1999- ),女,安徽凤阳人,中国矿业大学硕士生,主要研究方向为室内定位
    李航(1985- ),男,河北承德人,博士,深圳市大数据研究院副研究员,主要研究方向为定位技术和无线通信
    曹雯(1987- ),女,陕西宝鸡人,博士,长安大学副教授,主要研究方向为智能交通信息处理和多源信息融合等
    王洪梅(1983- ),女,山东诸城人,博士,中国矿业大学副教授,主要研究方向为无线通信
    熊海良(1981- ),男,湖南双峰人,博士,山东大学副教授,主要研究方向为导航与定位、智能感知、智能决策等
    李世银(1971- ),男,四川犍为人,博士,中国矿业大学教授、博士生导师,主要研究方向为煤矿信息化和移动目标定位
  • 基金资助:
    陕西省自然科学基础研究计划基金资助项目(2023-JC-YB-510);深地科学与工程云龙湖实验室基金资助项目(109023005);中央高校基本科研业务费专项资金资助项目(300102322103);山东省自然科学基金资助项目(ZR2022LZH005);青岛市科技惠民示范引导专项基金资助项目(22-3-7-CSPZ-2-nsh)

Research on geomagnetic indoor high-precision positioning algorithm based on generative model

Shuai MA1,2,3, Ke PEI3, Huayan QI3, Hang LI4, Wen CAO5, Hongmei WANG3, Hailiang XIONG6, Shiyin LI3   

  1. 1 Yunlong Lake Laboratory of Deep Underground Science and Engineering, Xuzhou 221116, China
    2 Peng Cheng Laboratory, Shenzhen 518172, China
    3 School of Information and Control Engineering, China University of Mining and Technology, Xuzhou 221116, China
    4 Shenzhen Research Institute of Big Data, Shenzhen 518172, China
    5 College of Electronics and Control Engineering, Chang’an University, Xi’an 710064, China
    6 School of Information Science and Engineering, Shandong University, Jinan 250100, China
  • Revised:2023-04-12 Online:2023-06-01 Published:2023-06-01
  • Supported by:
    The Natural Science Basic Research Program Foundation of Shaanxi Province(2023-JC-YB-510);Yunlong Lake Laboratory of Deep Underground Science and Engineering Project(109023005);The Fundamental Research Funds for the Central Universities(300102322103);The Natural Science Foundation of Shandong Province(ZR2022LZH005);Qingdao Science and Technology Demonstration and Guidance Project for Benefiting the People(22-3-7-CSPZ-2-nsh)

摘要:

针对目前构建精细的地磁指纹库需要耗费大量人力成本的瓶颈,提出了条件变分自动编码器和条件对抗生成网络2种生成模型,能够在收集少量数据样本的基础上,对给定位置进行伪标签指纹的生成。同时,针对单点地磁指纹定位精度低的问题,设计了一种基于注意力机制的卷积神经网络-门控循环单元的地磁序列定位算法,能够有效利用指纹的空间和时间特性,实现精准定位。此外,还设计并搭建了实时、便携的移动端数据采集和定位系统。通过实际测试表明,利用所提模型可有效构建可用的地磁指纹库,所提算法平均误差可达0.16 m。

关键词: 深度学习, 地磁定位, 生成模型, 地磁序列

Abstract:

Aiming at the current bottleneck of constructing a fine geomagnetic fingerprint library that required a lot of labor costs, two generative models called the conditional variational autoencoder and the conditional confrontational generative network were proposed, which could collect a small number of data samples for a given location, and generate pseudo-label fingerprints.At the same time, in order to solve the problem of low positioning accuracy of single-point geomagnetic fingerprints, a geomagnetic sequence positioning algorithm based on attention mechanism of convolutional neural network-gated recurrent unit was designed, which could effectively use the spatial and temporal characteristics of fingerprints to achieve precise positioning.In addition, a real-time, portable mobile terminal data collection and positioning system was also designed and built.The actual test shows that the proposed model can effectively construct the available geomagnetic fingerprint database, and the average error of the proposed algorithm can reach 0.16 m.

Key words: deep learning, geomagnetic positioning, generative model, geomagnetic sequence

中图分类号: 

No Suggested Reading articles found!