A coal flow measurement method for belt conveyors in coal mines based on multi-scale point cloud strategy

Abstract To address the challenge of low accuracy in coal flow detection for coal mine belt conveyors, this paper proposes a precise coal flow measurement method for belt conveyors in coal mines based on multi-scale 3D point cloud data processing. Firstly, both density-based adaptive statistical filtering and density-based automatic pass-through filtering are proposed to effectively remove irrelevant points from the point cloud and autonomously extract the target coal flow points. To reduce data processing complexity, an octree voxel filtering based on KD-Tree is used to downsample the coal flow point cloud. Then, the point cloud is projected onto a two-dimensional plane. Based on the positions of the marked pixel points, the coordinates of the filling points on the upper and lower surfaces of the coal flow point cloud are calculated using K-nearest neighbor regression interpolation and global interpolation, respectively. Finally, a center-of-mass-based tetrahedral hierarchical accumulation method is proposed to compute the point cloud volume by summing the tetrahedral volumes within all grid units of the point cloud. Experimental results show that within the belt speed range of 0–3 m s −1 , the average accuracy of coal flow rate calculation in 60 experiments is 97.51%, with a maximum error of 4.76% and an average computation time of 1.62 s, meeting the requirements for coal flow rate measurement in coal mine belt conveyors.