Research On Error Model Construction And Quality Evaluation Of Underground Mobile LiDAR Data

With the expansion of human activities in economic development,the use of underground space is becoming more and more frequent.However,the problems of small underground environment space and insufficient light lead to the development of data acquisition methods of underground 3D space information is still relatively slow.The spatial data acquisition of mobile LiDAR system has the advantages of fast speed and high accuracy,but in the underground environment without GNSS signal,mobile LiDAR system cannot acquire high-precision point cloud data for a long time.The key of high accuracy acquisition of 3-D point cloud information in underground space is to restrict the drift of integrated navigation error without GNSS.In this paper,the problem of obtaining high-precision point cloud information for a long time in the absence of GNSS such as underground is studied,and the methods described in this paper are comprehensively validated and analyzed.Specific research work is as follows:(1)Data fusion model reasoning of mobile LiDAR system.On the basis of briefly introducing the composition and structure of current mobile LiDAR system,the coordinate system involved in the acquisition process of three-dimensional point cloud data in mobile LiDAR system is described,and the geometric positioning model in the process of multi-sensor fusion solution of mobile LiDAR system is deduced.(2)Precision center positioning of graphical cooperative target.According to the different characteristics of laser beam echo intensity of different materials,a circular graphic cooperative target is used to provide a space-time reference for mobile LiDAR system.A weighted iterative M-estimation method based on Denmark method is proposed to locate the target center.Compared with the least square method,the RANSAC circle fitting method and the gravity center method,the results show that the method has the characteristics of high positioning accuracy and fast speed.Circular cooperative target can assist mobile LiDAR system to carry out high-precision space-time reference.(3)Error source analysis and error constraint model construction.The error sources of mobile LiDAR system are analyzed.It is shown that the main error sources of mobileLiDAR system in underground space are integrated navigation error.The Kalman filter method commonly used in current integrated navigation is introduced,and the trajectory acquisition model of GNSS/INS/cooperative target combination under the constraint of graphical target is deduced.(4)Evaluation of quality of point cloud data.On the basis of determining the selection criteria of point cloud data quality evaluation indicators,according to the characteristics of spatial data evaluation,point cloud gross error rate,point cloud density,signal-to-noise ratio of echo intensity are taken as the evaluation indicators of point cloud data quality,and elevation error and plane error are taken as accuracy indicators.(5)Experimental verification and result analysis.According to the characteristics of underground non-GNSS signal,the experimental environment is through the outdoor road.On the basis of obtaining good data of GNSS signal,the GNSS observation data in the middle period is deleted to simulate the GNSS state of underground space for data calculation.Then,the unconstrained,25 m,50m,100 m and 200 m constraints are set to verify the effectiveness of these constraints.The improvement of integrated navigation accuracy by different target intervals is analyzed,and the quality of point cloud data acquired by mobile LiDAR system is evaluated.The results show that the acquisition efficiency and accuracy are better when 50 m interval is set.