Phosphate Mine Roadway Environment Research On Unmanned Transporter’s Essential Technologies

Technology advancement has caused a progressive movement in the direction of intelligent and automated mining production in recent years.Unmanned transport vehicles in mining tunnels are one of the research hotspots,garnering considerable attention from the mining industry,since the country has placed increasing priority on the creation of smart mines.In order to complete the entire business process of underground loading,transporting,and unloading,unmanned transportation in tunnels is primarily intended to replace manual driving.This helps to address problems with mine safety production,occupational health and safety,transportation efficiency,and operating costs.Phosphate deposits are plentiful in Yichang,one of the major phosphate production hubs in the Yangtze River Basin.This article performs study on the primary autonomous navigation technologies for unmanned transport vehicles in the phosphate tunnel environment based on the local phosphate mining environment,specifically from the following aspects:(1)Extracting the boundary of the phosphate mine tunnel based on the laser radar point cloud data.A Mid-70 laser radar algorithm platform was built to scan typical phosphate mine tunnels and obtain the point cloud data of the tunnel.In order to reduce the influence of non-tunnel structural point clouds,a planar projection method was first used to obtain a two-dimensional planar point cloud map of the tunnel,and then point cloud filtering was carried out to reduce the data amount.Common boundary extraction algorithms were analyzed,and the improved Alpha-shapes algorithm was selected to extract the tunnel boundary based on the characteristics of the tunnel planar point cloud map,which improved the efficiency of boundary extraction.(2)Path planning based on the centerline of the tunnel.In order to solve the problem of unordered indexing of boundary points in the tunnel,a boundary point convex hull was created based on the direction of the tunnel boundary,and the boundary points were indexed in order.The Delaunay triangulation algorithm was analyzed,and a constrained Delaunay triangulation network was generated inside the boundary point convex hull using the incremental insertion method,and the triangle area was calculated.An area threshold was designed to filter out small triangles,and the final triangulation network used to construct the tunnel centerline was obtained.The Voronoi diagram method was analyzed,and a non-boundary Voronoi diagram was drawn on the triangulation network using an improved algorithm to construct the tunnel centerline.The improved Dijkstra algorithm was used to search for a path on the tunnel centerline,and the B-spline curve was used to smooth the path,completing the global path planning for unmanned transport vehicles in the tunnel.(3)Designing obstacle avoidance strategies in the mine tunnel based on the obstacle point cloud detection results.The interested area’s obstacle point cloud was obtained by using the laser radar scanning.The density and distance-based clustering algorithms were analyzed,and improvements were made according to the distribution characteristics of the tunnel obstacles to complete the obstacle clustering.The OBB bounding box was used to envelop the obstacle point cloud,and the size information of the obstacles was obtained.Combining with the tunnel’s enclosed characteristics,the obstacle occupancy was calculated,and the obstacle type was determined based on the velocity characteristics of the objects in the tunnel.The traffic environment of the tunnel was analyzed,and the obstacle avoidance strategy for typical tunnel scenarios was designed based on the detection results.(4)Joint simulation experiment and intelligent vehicle validation.Based on the Carsim simulation software,a vehicle and typical tunnel scene model are built,and data is transmitted to Matlab/Simulink.Through PI control,the adjustment information is returned to Carsim,forming a closed-loop system to complete the joint simulation.The simulation results show that the unmanned vehicle can accurately avoid obstacles according to the corresponding obstacle avoidance strategy based on the detected obstacles during the driving process along the centerline path.At the same time,the centerline path planning method was verified using an intelligent vehicle in a laboratory corridor simulating a phosphate mine tunnel.