Study On Deformation Monitoring Of Curved Tunnel Section Based On Land Based LiDAR

In order to identify risks and early warning during tunnel construction and operation,deformation monitoring must be carried out.Although the precision of the traditional method is high,it can reach the mm level,but there are also many shortcomings: the discrete point of the discrete section,the precision of the precision affected by the light,the limited number of the measured section and so on.This paper introduces a three dimensional laser scanning technology,the land based LiDAR technology,and selects an experimental tunnel near the Three Gorges as an experimental site.According to the inner diameter of the tunnel and the point cloud precision required for the project,the distance between the stations and the scanning resolution are determined,and a detailed scanning scheme is made,and two point cloud data are collected one year.This paper analyzes the factors that affect the accuracy of tunnel point cloud registration from three aspects,such as the type of target,the number of targets and the introduction of control points.After projection of three-dimensional tunnel point cloud data by bidirectional projection method,the projection boundary point is extracted by the segmentation method and the improved RANSAC algorithm is used to fit the boundary curve.The axis of the curve tunnel is extracted.In order to extract the point cloud of tunnel section quickly and accurately,the concept of point buffer is put forward.In order to ensure that the extracted section has enough point cloud,the projection surface buffer area is set on the 3D point cloud.The thickness of the surface buffer directly affects the accuracy of the section deformation analysis,and the concept of the optimum surface buffer thickness is put forward and the optimal thickness is determined.The MDP algorithm is used to calculate the deformation of the tunnel section at any position and at any time.But the algorithm is used to estimate the noise of the deformation signal on the tunnel surface.In this paper,the covariance function is used to identify the real deformation signal to reduce the deformation error.The main conclusions of this paper are as follows:(1)4 spherical targets are used to optimize the registration of point cloud data between stations.(2)the improved RANSAC algorithm successfully extracted the central axis of the curved tunnel.(3)under the premise that the best buffer thickness is 4mm,the point cloud of the curved tunnel is extracted.(4)compared with total station,the accuracy of cross section deformation extracted in this paper is within 3 mm.(5)the deformation accuracy of tunnel section calculated by MDP algorithm is ± 3 mm,which is compared with the accuracy of ±10 mm,which is calculated by the common point cloud processing software Geomagic Qualify and CloudCompare.The feasibility of the MDP algorithm in the analysis and application of the curve tunnel deformation is verified.