Research On 3D Mapping Method Based On LiDAR For Underground Roadway

In recent years,the implementation of the national energy security strategy has been upgraded,the integration of information technology and coal industry has accelerated,and the intelligent transformation and upgrading of the coal industry is imminent.For this reason,this paper takes the key equipment of coal mine roadway-cantilever roadheader as the research object,aiming to realize its autonomous positioning and autonomous cutting in the roadway,and realizes a three-dimensional mapping method of coal mine underground roadway based on single-line laser radar.,The main work is as follows:(1)According to the characteristics of underground tunnels in coal mines and the intelligentized requirements of cantilever roadheaders,a three-dimensional laser scanning system have built,the factors affecting the accuracy of the system have been analyzed,and the errors of the three-dimensional points scanned by the system were tested experimentally.The built 3D laser scanning system is composed of a 2D laser radar and a pitching rotary table.And the pitch scan angle is greater than ±90°.The main factors affecting the accuracy of the system are the ranging accuracy of lidar,the repeated positioning accuracy of the pitching rotary table and the mechanical installation error.In the error test experiment,when the 3D laser scanning system scans the plane at 10 m,the distance between the obtained 3D point and the fitting plane is not more than 0.11 m.(2)According to the characteristics of the coal mine roadway point cloud data and the roadway mapping requirements,the segmentation of point cloud data of coal mine roadway working face has been achieved by using cloth filtering algorithm.Cloth filtering does not rely on geometric features for segmentation,and has a good effect on the segmentation of working face point clouds similar to the surrounding environment features.In the simulation experiment,the error of class I is less than 8.22%,the error of class II is less than 0.36%,and the error rate of point cloud delineation is less than0.58%.(3)The positioning error of roadheader based on UWB-IMU has been analyzed,and LiDAR SLAM has been realized based on UWB-IMU positioning.In the process of implementing LiDAR SLAM,the matching search space is determined according to the UWB-IMU positioning error,and the ICP point cloud matching algorithm is used to realize the splicing of adjacent frame point clouds,thereby constructing a roadway map and improving the positioning accuracy of the roadheader.The simulation results show that the vertical positioning accuracy is increased to within 0.075 m,and the heading angle accuracy has increased to within 0.078 rad.In this paper,a 3D laser scanning system has constructed to obtain the basis of roadway point cloud data.Based on the cloth filtering algorithm,the point cloud data of the coal mine roadway face is segmented.Based on the UWB-IMU positioning,LiDAR SLAM is realized,the roadway map is constructed,and the roadheader is improved.positioning accuracy.