A deformation monitoring method for coal mine roadway based on 3D laser scanning

Abstract

The traditional monitoring methods for coal mine roadway deformation have problems such as incomplete data collection, non intuitive data format, poor precision, and inability to achieve continuous monitoring of the entire roadway deformation. In order to solve the above problems, a deformation monitoring method for coal mine roadway based on 3D laser scanning is proposed. Firstly, the method uses 3D laser scanning technology to obtain real 3D point cloud data of coal mine roadways. Secondly, the deep learning model VoxelNet is used to detect and denoise 3D laser scanning data, converting unordered point cloud data into high-dimensional feature data. The Alphashape algorithm is used to fit the discrete points of the extracted roadway cross-section. The multi-dimensional difference calculation based on the difference method is used to obtain specific data of roadway deformation, achieving full coverage of roadway deformation monitoring in the mining area. The 3D laser scanning technology is applied to deformation monitoring of the 30507 working face in Tashan Coal Mine. The cross-sectional analysis and 3D overall analysis are conducted on the 3D point cloud data of the roadway. The analysis results indicate that the main deformation in the area can be directly observed through the relative deviation of the section contour of the second stage roadway. If the upper contour deviates inward, the roof will collapse. If the lower contour deviates outward, the floor will bulge. As the distance between the measuring point and the working face gets closer, the color of the attached color model leans towards red and blue. The darker the color, the greater the deformation of the roadway.

Keywords

• deformation monitoring of coal mine roadway
• 3d laser scanning
• point cloud data
• cross-sectional profile of the roadway
• point cloud denoising
• point cloud rotation