Application Research Of Laser SLAM Mine Underground Space Measurement

With the continuous improvement of mine safety indicators and the attention to the investigation and management of mined out areas,the single underground measurement system can not meet the needs of information construction.However,the virtual mine model established by traditional surveying and mapping technology can not reflect the original landform of the whole mine completely and accurately due to its small number of measurement points,so it is difficult for the state to automatically mine The demand of mining and intelligent mining.After years of development,the vehicle mobile measurement system has been quite perfect.The system integrates inertial navigation technology,laser scanning,panoramic camera and other technologies,which can obtain the point cloud and image information of surrounding ground features in the process of rapid movement,but it can not go deep into the building to scan the indoor or underground space and other narrow special environment.Because the system relies too much on GNSS/INS navigation information,there is no GPS signal in the underground space environment of mine GNSS/INS integrated navigation can not provide high-precision positioning information,and can not rely on mobile measurement to obtain high-precision map.Simultaneous location and mapping(SLAM)is a relative positioning method.Its core is to use lidar to scan the surrounding environment information,and calculate the carrier's POS by point cloud feature extraction,point cloud registration and other algorithms.In recent years,the backpack mobile measurement systems based on laser slam emerge in endlessly.Aiming at the problem of high-precision composition of the system in the harsh environment with highly similar surrounding environment and no obvious characteristics,this paper studies the relative spatial position calibration of the sensor composed of hardware and the improvement of the algorithm(1)Calibration of relative spatial position of sensor.The calibration results of sensors directly affect the mapping accuracy of mobile measurement system.In this paper,the relative spatial relationship between the carrier and the laser scanner and between the two laser scanners are calibrated.The calibration accuracy is 0.051cm,and the calibration accuracy is less than 2cm by plane model conversion method.(2)Research on depth first branch and bound algorithm.The node selection,branch rules and upper bound calculation of the algorithm are studied in detail.The simulation results show that the improved algorithm can shorten the drawing time.(3)Research on SPA algorithm.The improved spa algorithm based on LM framework is used to optimize the continuous LM algorithm,reduce the dependence of the original LM algorithm on the initial value,and achieve efficient calculation through a pre calculated ordered data structure.(4)The improvement of loopback mechanism in loopback detection.Cartographer algorithm is prone to loopback errors in harsh corridors or narrow roadways with long distance,equal width and no obvious features around,which eventually leads to the decline of composition accuracy.In this paper,through the research and design,the lazy decision decision decision is added to the loop detection.Through the simulation test of multiple sets of data,the algorithm can effectively improve the quality of loop detection and the accuracy and robustness of drawing.(5)A 3D laser slam measurement system with backpack is built.Through the actual measurement experiments of large underground parking lot without GPS signal and mine roadway and shaft with harsh mapping environment,the measured point cloud image is highly consistent with the real scene.Finally,the experimental results are analyzed from the accuracy of mapping and positioning,algorithm robustness,computational complexity and real-time performance.The relative mapping accuracy of the mobile measurement system in this paper is 5 cm left Right,it has high robustness and low algorithm complexity in harsh environment,which can meet the measurement task of underground space in mine.