3D Reconstruction And Deformation Analysis Of Coal Mine Tunneling Based On Laser Point Cloud

Coal as main energy at a time in our country,is the nation’s economic lifeline,but coal production environment is extremely bad,roadway deformation will not only broken the production environment,caused by underground traffic paralysis,will directly affect the mine production equipment and the safety of his staff,production produce a great threat to the safety of coal mine.However,the existing roadway deformation monitoring is mostly contact measurement method,which not only consumes a lot of manpower and material resources,but also is difficult to achieve the overall measurement range,which is contrary to the trend of intelligent coal mine.In this paper,aiming at the monitoring and early warning of roadway deformation in coal mine,taking driving face as an example,a set of 3d tunnel reconstruction experiment system based on laser point cloud was developed,and the experimental verification was carried out in the simulated roadway.The 3d tunnel point cloud map was established,and the deformation of roadway was analyzed and calculated according to the reconstruction results.The main research results are as follows:(1)A simple 3D scanning device was designed based on multi-beam lidar and high-precision rotating pylon.Modbus communication and lidar communication of rotating pylon were written by C++.Make it realize automatic omni-directional(360°× 360°)scanning and point cloud collection,and establish3 d laser point cloud map of the current collection location through point cloud registration method;(2)Simulate and analyze the distribution of the projection point cloud in the orthogonal plane of the scanning system under ideal conditions by means of simulation in MATLAB,and feedback and modify the actual parameters of the single-station scanning system according to the simulation results;(3)The integrated navigation system(strapdown inertial navigation system and total station)is added to provide the position(translation vector t)and attitude(rotation matrix R)information between each two scanning sites,so as to realize the splicing and matching of point cloud data obtained at different scanning locations;(4)The method of KNN+ least square surface fitting is used to calculate the matching error of different stations in the data of the same period,and the calculated average distance error is in the centimeter level.And by comparing the actual measured observation value with the estimated value obtained from the point cloud measurement on the software,the scanning precision of the scanning equipment is oriented,the average error is less than 5cm,and the error rate is less than 1%.(5)The point cloud data is preprocessed,and the projection method is used to calculate the central axis of the THREE-DIMENSIONAL point cloud.First,the fitting curve of the outer contour points of the projection point cloud in the XOY plane is calculated,and the polynomial function is determined to be used for fitting.By calculating the values of RSME and R-square,the order of the polynomial is determined as: 3.Then,the fitting curve of the outer contour points of the projection point cloud on the YOZ plane was calculated.According to the fitting curve of the outer contour points,the fitting curve of the entire projection point cloud was calculated,and the fitting curve of the three-dimensional point cloud was finally obtained,namely the central axis.(6)According to the laser point cloud,the deformation of the roadway is analyzed.The data of the two phase point clouds at the same position on the sectional plane are obtained through the central axis.The distance between points is calculated by selecting points in adjacent areas and averaging,so as to calculate the displacement of different monitoring positions.