CN-117197205-B - Non-rigid registration method for tunnel deformation steel arch

CN117197205BCN 117197205 BCN117197205 BCN 117197205BCN-117197205-B

Abstract

The invention provides a non-rigid registration method for a tunnel deformation steel arch, which belongs to the crossing field of a point cloud processing technology and tunnel engineering, and mainly comprises 3 stages, namely (1) point cloud pretreatment, (2) extraction of a steel arch space trend characteristic curve, and (3) non-rigid registration of the point cloud based on a space curve. According to the distribution rule of the polar diameters of the neighboring points, the separation of the rock wall part, the non-rock wall part and the steel arch part from noise in the original scanning point cloud is sequentially realized by an iteration method, then the spatial trend characteristic curve of the steel arch point cloud obtained after the pretreatment of the steel arch design model and the point cloud is extracted respectively, and finally the non-rigid registration is carried out on the design model and the steel arch point cloud based on the characteristic curve, so that the recovery of the steel arch in the tunnel is realized. The invention aims at the forced torsion deformation steel arch in the tunnel, can realize the efficient and accurate steel arch recovery, provides technical support for the information perception of intelligent equipment on the tunnel rock wall support, and provides parameter basis for the realization of automatic wet spraying operation.

Inventors

XIAO JUN
ZHANG QIN
LIU LUPENG
WANG YUNBIAO
YU DONGBO

Assignees

中国科学院大学

Dates

Publication Date: 20260512
Application Date: 20230919

Claims (7)

1. A point cloud pretreatment method for a tunnel deformation steel arch, characterized by comprising the following steps: step 1, preprocessing point cloud, namely separating rock wall and non-rock wall in original scanning point cloud and a steel arch from a noise part; step 2, extracting the curve of the design model, namely neglecting the local characteristics of the steel arch design model, only keeping the overall trend of the steel arch design model, and extracting and constructing the space trend characteristic curve of the steel arch design model to approximately describe the steel arch design model; step 3, extracting a curve of the steel arch point cloud, namely ignoring local characteristics of the steel arch point cloud, only keeping overall trend of the local characteristics, and extracting and constructing a space trend characteristic curve of the steel arch point cloud to approximately describe the steel arch point cloud; Step 4, point cloud registration based on space curves, namely further realizing non-rigid registration of the steel arch design model and the steel arch point cloud by matching the space trend characteristic curves output in the step 2 and the step 3; Said step 1 further comprises the steps of: step 1.1, coordinate transformation and coordinate system transformation, namely sequentially carrying out coordinate translation, coordinate rotation and coordinate system transformation on the original scanning point cloud to obtain a scanning point cloud after coordinate system transformation; step 1.2, denoising the blasting point cloud, namely dividing the point cloud output in the step 1.1 into two parts of real rock wall and steel arch noise; step 1.3, denoising the steel arch point cloud, namely dividing the steel arch noise point cloud output in the step 1.2 into a steel arch and noise parts; Said step 1.2 further comprises the steps of: firstly, constructing a virtual rock wall point cloud, namely weighting and calculating the virtual rock wall polar diameter corresponding to the current point by each point of the point cloud output in the step 1.1 according to the neighbor point information of each point by the neighbor point Specifically, the method comprises, in particular, , Wherein the method comprises the steps of For the number of neighbor points, Is taken as a point Adjacent to the point In particular, the weight of the (c) in the (c) is determined, , Wherein the first part of the right-hand multiplication of the equal sign depends on the point Adjacent to the point In a direction perpendicular to The Euclidean distance of the projection on the plane of the direction, the second part being dependent on the point Adjacent to the point At the polar diameter Difference in direction; Calculating the polar diameter variation, namely calculating the polar diameter of the virtual rock wall for each point of the point cloud output in the step 1.1 A pole diameter difference value with the pole diameter r value; And finally, iterating to remove the non-rock wall point cloud, namely adopting an iteration method to enable the point cloud output in the step 1.1 to be according to the polar diameter variation Performing iterative denoising on the numerical value of (1); said step 1.3 further comprises the steps of: Firstly, mapping point clouds, namely, the point clouds output in the step 1.2 are mapped from a cylindrical coordinate system Mapping to a three-dimensional rectangular coordinate system Lower two-dimensional plane In (a) and (b); grouping the point clouds, namely mapping the point clouds to a plane point set according to polar angles Equidistant segmentation is carried out on the values to obtain a plurality of groups of local point clouds; and finally, iteratively removing noise points, namely adopting an iteration method to remove sheet noise from each group of local point clouds according to the distribution density in the polar diameter r direction.
2. The point cloud pretreatment method for the tunnel deformation steel arch according to claim 1, wherein the method comprises the following steps: said step2 further comprises the steps of: Step 2.1, point cloud construction, namely reading a steel arch design point cloud file, and constructing a corresponding point cloud of a steel arch design model according to read data; step 2.2, segmenting point cloud, namely corresponding point cloud basis of a steel arch design model Equidistant segmentation is carried out on the values, each segment of point cloud is represented by a characteristic point, and the characteristic point is a node corresponding to the current segmentation; step 2.3, assigning node coordinates, namely assigning the x coordinate of the node to be zero, and assigning the node to be zero Assigning coordinates to all points contained in the segment corresponding to the node Assigning the r coordinate of the node as the minimum value of r of all points contained in the corresponding segment of the node; and 2.4, constructing a space curve of the design model, namely representing the space curve by using the extracted space point sequence.
3. The point cloud pretreatment method for the tunnel deformation steel arch according to claim 1, wherein the method comprises the following steps: said step 3 further comprises the steps of: Step 3.1, mapping point cloud, namely mapping the point cloud of the steel arch into a two-dimensional point set on a plane; step 3.2, constructing a plane curve, namely performing two-dimensional plane curve fitting on the two-dimensional point sets, and respectively solving a plane curve function corresponding to each distorted steel arch frame mapped to the two-dimensional point sets ; Step 3.3, constructing a space curve of the point cloud of the steel arch, wherein the space curve is combined with the plane curve function output in the step 2.2 The deformation result of each twisted steel arch frame in the radial direction And solving.
4. A point cloud pretreatment method for a tunnel deformation steel arch according to claim 3, characterized in that: said step 3.2 further comprises the steps of: Firstly, carrying out point set segmentation, namely segmenting all points in the two-dimensional point set according to the following conditions Equidistant segmentation is carried out on the size of each segment of point set, wherein the distribution of each segment of point set on a two-dimensional plane is a plurality of approximately parallel line segments, and each line segment corresponds to different steel arches; Secondly, extracting two-dimensional line segments, namely respectively extracting plane line segments from each segment of point set; Finally, carrying out plane curve construction, namely carrying out one-to-one pairing on a plurality of plane line segments corresponding to two adjacent point sets according to the size of an x coordinate value, sequentially ending and connecting the paired adjacent line segments to form a plane curve, and completing the plane curve function Is a structure of (a).
5. The point cloud pretreatment method for the tunnel deformation steel arch according to claim 4, wherein the point cloud pretreatment method comprises the following steps: The two-dimensional line segment extraction is based on Hough transformation two-dimensional straight line identification, and further comprises the following steps: Firstly, carrying out linear extraction, namely respectively carrying out two-dimensional linear extraction based on Hough transformation on each section of point set, wherein the number of linear extraction is the number of steel arches identified by the current section; secondly, determining the number of the steel arches, wherein the number of the identified steel arches in each section of point set is not necessarily the same, and the number of the identified steel arches is required to be unified; finally, line segment interception is carried out, namely all extracted straight lines are grouped at present And taking the boundary value as a reference, and intercepting the line segment.
6. A point cloud pretreatment method for a tunnel deformation steel arch according to claim 3, characterized in that: said step 3.3 further comprises the steps of: first, for the plane curve function According to Equidistant sampling is carried out on the size; Second, each sampling point is traversed Searching the sampling points in a two-dimensional point set on an ideal plane Is a plurality of adjacent points; Then, searching the polar diameter value of each point in the plurality of adjacent points under the original cylindrical coordinate system before mapping according to the point cloud index, and assigning the minimum value in the polar diameter values of the plurality of adjacent points under the original cylindrical coordinate system to Completing the deformation result of the polar radial direction Is solved; finally, combining the plane curve functions The deformation result in the radial direction By traversing And finishing the construction of the space curve of the steel arch point cloud.
7. The point cloud pretreatment method for the tunnel deformation steel arch according to claim 1, wherein the method comprises the following steps: Said step4 further comprises the steps of: Step 4.1, the spatial trend characteristic curve of the steel arch point cloud and the spatial trend characteristic curve of the steel arch design model are combined to form the steel arch design model Representing the independent variable; step 4.2, calculating the difference value between the spatial trend characteristic curve of the steel arch point cloud and the spatial trend characteristic curve of the steel arch design model to obtain the variation of curve coordinates And (3) with ; Step 4.3, traversing each point in the corresponding point cloud of the steel arch design model Superimposing the points by vectors Performing displacement; and 4.4, storing the result of each point displacement in the corresponding point cloud of the steel arch design model as a new point cloud and outputting the new point cloud.

Description

Non-rigid registration method for tunnel deformation steel arch Technical Field The invention belongs to the crossing field of a point cloud processing technology and tunnel engineering, and particularly relates to a non-rigid registration method for a tunnel deformation steel arch. Background In the construction of tunnel engineering, tunnel excavation is usually realized by a drilling and blasting method, a shield method or a combination of the two methods. Because the shield method has the defects of high requirements on geological environment, poor adaptability to section change and the like, the method generally does not have flexibility and adaptability of construction, and therefore, when tunnel construction faces extreme stratum structures such as frequent lithology change, uneven hardness, multiple faults and the like, the construction needs to be additionally dependent on a drilling and blasting method. After the tunnel is excavated by the drilling and blasting method, reinforced concrete lining is paved, and the initial support of the tunnel is realized by spraying concrete and supporting by matching with a steel arch frame, so that surrounding rock collapse is prevented. In the conventional drilling and blasting method construction, the concrete wet-spraying operation requires the construction personnel to operate on site. And severe environmental factors such as rock burst, large deformation, high temperature, damp and heat, improper ventilation and the like in the tunnel environment can influence the construction process, so that the problems of low construction efficiency, poor safety, difficult guarantee of the operation quality and the like are caused. By developing novel automatic construction equipment, the defects of insufficient mechanization and the like of the traditional drilling and blasting method are overcome, the labor intensity of operators is reduced, the life safety of the operators is guaranteed, the safe, efficient, automatic and mechanized tunnel construction under various geological problems is realized, and the method has important significance for the technical development of tunnel engineering. In the automatic wet spraying operation, the region, which is shielded by the steel arch, of the tunnel rock wall cannot be directly sprayed. The gap between the steel arch and the rock wall is filled in a manner that the mechanical arm bypasses the steel arch in the tunnel by planning a path of the mechanical arm of the wet spraying table machine. In order to guide the path planning of the mechanical arm, steel arch recovery is required to be carried out on the steel arch point cloud extracted from the blasting point cloud, the perception of steel arch position information by the intelligent equipment is realized, the intelligent equipment is used for distinguishing the steel arch position to provide parameter support for the path planning of the mechanical arm of the wet spraying platform machine, and finally, the intelligent equipment replaces site constructors to guide the wet spraying platform machine to carry out automatic wet spraying operation. Meanwhile, the steel arch recovery result can be used for removing the volume occupied by the steel arch part in the block to be filled in the concrete injection amount calculation result so as to improve the accuracy of concrete injection amount calculation. Due to the limitation of the precision of the laser scanning equipment in the actual engineering, the information acquisition of the real steel arch is insufficient, the local characteristics of the scanning point cloud cannot be fully reflected in the original blasting point cloud file derived from the equipment, and the recovery and reconstruction of the steel arch from the real steel arch point cloud cannot be directly realized. By using a non-rigid registration technology for the steel arch point cloud, the steel arch design model is matched into the reconstructed rock wall model according to the registration result of the denoised steel arch design model and the steel arch point cloud, so that the effective recovery of the distorted steel arch in the tunnel can be realized, and the steel arch recovery problem is solved. Disclosure of Invention The invention aims to provide a recovery method of a forced torsion deformation steel arch in a tunnel. In order to achieve the technical purpose, the technical scheme adopted by the invention comprises two parts of point cloud preprocessing and non-rigid registration. The point cloud preprocessing is a preprocessing part of the non-rigid registration method, and the non-rigid registration method is divided into two stages of extraction of a space trend characteristic curve of the steel arch and non-rigid registration of the point cloud based on the space curve. Embodiments provide a non-rigid registration method for a tunnel deformation steel arch. Step1, preprocessing point cloud, namely separating rock wall and non-rock wall in o