The Study Of Deformation Monitoring Based On Subway Shield Tunnel Point Clouds

The development of rail transit,especially subway,is an effective means to solve the difficulties of riding in big cities and alleviate road congestion.Subsequent subway tunnel safety issues are receiving increasing attention.How to efficiently and accurately monitor subway tunnel deformation is one of the hot and difficult issues in underground engineering research.Terrestrial laser scanning technology has been widely used in many fields because of its advantages of short data acquisition period,complete scanning scene,high accuracy,no contact,and no environmental restrictions.Applying it to subway engineering safety monitoring can provide important technical support for subway engineering safety monitoring and operation management.In this paper,the ground laser scanning technology is applied to the safety monitoring of shield subway tunnels.Based on the point cloud data characteristics of shield tunnels,the safety monitoring method based on point cloud data is studied.The main research contents and results of this article are as follows:(1)The research puts forward the layout of the ground laser scanning work reference point and measuring station of the shield tunnel,the scanning resolution and scanning quality settings,the point cloud data collection method,and analyzes the characteristics of the tunnel point cloud data.(2)Aiming at the problem of mixing straight and curved lines of shield tunnels,a hybrid linear segmentation method based on "spatial transformation" is proposed.Through space transformation,the hybrid linear segmentation problem of the tunnel axis is transformed into a single linear recognition problem in the tangent space,and the point cloud of straight and curved tunnels is accurately segmented.(3)The shield tunnel section extraction method based on point cloud data is studied.Based on the segmented shield tunnel point cloud data,a tunnel section extraction method based on local cylindrical surface fitting interpolation is proposed.First,construct the ray in the 360 ° direction with the axis as the center in the normal plane,then fit the tunnel point cloud within a certain range of each ray to a cylindrical surface separately,calculate the intersection point of the ray and the cylindrical surface as the tunnel cross-section point;use RANSAC The algorithm fits the cylindrical equation,which effectively suppresses the influence of noise,points absence and other point cloud defects on the section extraction.Because the cylindrical section equation that fits the geometry of the tunnel surface is used to construct the tunnel section,the extracted tunnel section is more accurate.(4)The method of expanding the three-dimensional tunnel point cloud into a twodimensional gray image through the circular cross-section is studied,and the sequence relationship between the cross-section and the cross-section of the tunnel point cloud is restored in the form of cross-section resampling.Based on this sequence relationship,the sequence relationship is effectively Expanding the TLS tunnel point cloud data on a twodimensional plane overcomes the problem of deformation of the curve segment tunnel point cloud after expansion,and realizes the detection of tunnel segment seepage and shedding on the two-dimensional reflection intensity image.(5)The segmentation method of the tunnel point cloud on the ring-wise and tube-wise scales is studied.This method uses(4)the method of expanding the tunnel point cloud to a two-dimensional image to innovatively expand the geometric characteristics of the tunnel point cloud to two-dimensional On the plane,on this basis,the seal groove point cloud is used as the dividing line between the tunnel rings to complete the division between the rings.Using the bolt hole distribution characteristics of the capping block,the position of the capping block is identified ring by ring,and then the capping block is used according to Based on the arrangement relationship of the tube segments in the ring,the location of other types of tube segments can be determined,so as to complete the segmentation on the segment scale,which can effectively identify the segment stagger phenomenon.(6)Take a subway tunnel in Nanjing and Changzhou as examples,collect the point cloud data of the tunnel on the spot,and conduct an experimental analysis.The test analysis results show that the tunnel section extraction method proposed in this paper can effectively identify the tunnel convergence.With the point cloud density of 6mm / 10 m,the section extraction accuracy reaches 1.4 mm.Compared with the design radius,the section extraction accuracy is 1.60 mm The point cloud segmentation method proposed in this paper is not affected by the noise and lack of the tunnel point cloud.It can also effectively segment the point cloud in the case of uneven point cloud density,and can effectively identify the segment disease and stagger.