Lidar based map construction fusion method for underground coal mine shaft bottom

Guanghui XUE; Ruixue LI; Zhenghao ZHANG; Shuang LIU; Jinbo WEI

doi:10.13199/j.cnki.cst.2022-1111

Meitan kexue jishu (Aug 2023)

Lidar based map construction fusion method for underground coal mine shaft bottom

Guanghui XUE,
Ruixue LI,
Zhenghao ZHANG,
Shuang LIU,
Jinbo WEI

Affiliations

Guanghui XUE: School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083, China
Ruixue LI: School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083, China
Zhenghao ZHANG: School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083, China
Shuang LIU: School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083, China
Jinbo WEI: School of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083, China

DOI: https://doi.org/10.13199/j.cnki.cst.2022-1111
Journal volume & issue: Vol. 51, no. 8
pp. 219 – 227

Abstract

Read online

Intellectualization of coal mine is the technical support for high-quality development of coal industry, and robot replacement of key posts is the development trend of realizing efficient mining of coal with few people and no people. Simultaneous localization and mapping (SLAM) is one of the key technologies for autonomous movement and navigation of coal mine robots. The environment of underground coal mine is a typical unstructured environment, with narrow space, complex and changeable structure and uneven lighting, posing a severe challenge to the realization of SLAM in the underground coal mine. The research status of the map construction of the underground coal mine is summarized. In view of the shortcomings of the loopback detection of the LeGO-LOAM algorithm, the SegMatch algorithm is used to improve the loopback detection module of the LeGO-LOAM, the ICP algorithm is used to optimize the global map, and an improved algorithm integrating LeGO-LOAM and SegMatch is proposed, and the principle and implementation of the algorithm are discussed. The underground simulation scene experiments of coal mine were carried out, the mapping effect and accuracy of SLAM algorithm before and after the improvement were compared and analyzed, and the results showed that the map loopback effect constructed by the improved algorithm was better, and the estimated trajectory was smoother and more accurate. The construction method of two-dimensional occupied grid map is studied aiming at the navigation requirements, and the accuracy of the grid map constructed by this method is verified through experiments. The results show that the grid map after effectively filtering outliers such as dynamic obstacles has a mapping accuracy of 0.01 m, and the required storage space is 3 orders of magnitude lower than that of the point cloud map. The research results are helpful to the realization of SLAM and real-time positioning and autonomous navigation of the coal mine robot under the unstructured environment of the underground coal mine.

Published in Meitan kexue jishu

ISSN: 0253-2336 (Print)
Publisher: Editorial Department of Coal Science and Technology
Country of publisher: China
LCC subjects: Technology: Mining engineering. Metallurgy
Website: http://www.mtkxjs.com.cn/

About the journal

Abstract

Keywords