Research On Coal-rock Recognition Technology Of Fully-mechanized Coal Mining Face Based On Laser Scanning

The unmanned or less-humanized of fully-mechanized coal mining face is of great significance.As one of the core equipment of fully-mechanized coal mining face,the intelligent level of shearer is the key factor affecting the safety mining and production efficiency.The coal-rock recognition is the core technology to realize the intelligent mining of shearer.This thesis considers the laser point cloud data of coal-rock cutting surface in fully-mechanized mining face as the research object.The methods and technologies focus on the simplification and segmentation of coal-rock laser point cloud data.The accurate recognition of coal-rock is realized The main work in this thesis can be summarized as follow:(1)The fully-mechanized mining three machines cooperative work process is analysed.Combined the laser scanning technology with the actual needs of fully-mechanized coal mining face,the framework of the coal-rock recognition system for fully-mechanized mining face is established,and the process of coal-rock recognition based on laser scanning is designed.(2)A simplification method of coal-rock laser point cloud data based on feature points preserving is designed.The voxel-grid simplification method is used to preliminary simplify the point cloud data.The K-Means clustering method based on octree is designed.The k-d tree method and the least square method are used to obtain the feature points in each cluster.The feature points are fused with the preliminary simplified point cloud data to realize the simplification of coal-rock point cloud data.(3)A segmentation algorithm combining traditional region growing algorithm with point cloud intensity information is designed.The improved ant colony algorithm is used to optimize the two-dimensional OTSU(IACO-OTSU)method to obtain the optimal intensity threshold.The optimal intensity threshold is used to supplement the growing rule of the traditional region growing algorithm.The segmentation of laser point cloud data of coal-rock is completed,and the accurate recognition of coal-rock is realized.(4)The experimental system of coal-rock recognition based on laser scanning is designed and built.The experiment is carried out in Jiangsu Province and Education Ministry Co-sponsored Collaborative Innovation Center of Intelligent Mining Equipment and 13200 fully-mechanized coal mining face of Gengcun Coal Mine of Henan Dayou Energy Co.,Ltd.The experimental results show that the coal-rock laser point cloud data simplification method proposed in this thesis has better simplification results,and the surface deviation is small after reconstruction.The improved region growing algorithm based on IACO-OTSU can accurately segment the coal-rock laser point cloud data,and the accuracy of coal-rock recognition is more than 90%.The feasibility of the proposed method is verified.In this thesis,there are 41 figures,21 tables and 83 references.