CN-122021070-A - Segment deformation prediction method based on comparison of scanning data and simulation data

CN122021070ACN 122021070 ACN122021070 ACN 122021070ACN-122021070-A

Abstract

The invention relates to the technical field of bridge engineering, and particularly discloses a segment deformation prediction method based on comparison of scanning data and simulation data, which comprises the steps of S1, setting control points on the basis of a patch model of a bridge segment, carrying out deformation analysis on the patch model according to set process steps to obtain simulation deformation of the control points in corresponding process steps, S2, obtaining actual measurement deformation of the control points and three-dimensional coordinate data of the control points based on point cloud data of the bridge segment in the corresponding process steps, S3, repeating the steps of S1 to S2 to obtain a plurality of simulation deformation, actual measurement deformation and three-dimensional coordinate data corresponding to the control points to generate a data set, S4, inputting the data set into a built neural network model for training to obtain a trained deformation prediction model, and S5, carrying out deformation prediction on the bridge segment to be predicted according to the deformation prediction model. According to the invention, the simulation result is corrected through the actual measurement data, so that the accuracy of the simulation result is improved.

Inventors

ZHOU YONGJUN
LI KE
ZOU YU
FENG XIAOYANG
YANG ZHI
ZHANG TIANYI
Lu Daibin
YU WEI
SUN LIXIONG
He Enhuai
HUANG BING
LU WEI
YU ZHIBING
LIU XIAO
ZHAO LUXIN

Assignees

四川省钢构智造有限公司
四川路桥建设集团股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (9)

1. A segment deformation prediction method based on comparison of scan data and simulation data, comprising: s1, setting control points based on a surface patch model of a bridge segment, and carrying out deformation analysis on the surface patch model according to set process steps to obtain simulation deformation of the control points in the corresponding process steps; s2, acquiring the actually measured deformation of the control point and the three-dimensional coordinate data of the control point based on the point cloud data of the bridge segment in the corresponding process step; s3, repeating the steps S1 to S2, and obtaining a plurality of simulation deformation amounts, actual measurement deformation amounts and three-dimensional coordinate data corresponding to the control points to generate a data set; S4, inputting the data set into the constructed neural network model for training to obtain a trained deformation prediction model; and S5, carrying out deformation prediction on the bridge segment to be predicted according to the deformation prediction model.
2. The segment deformation prediction method based on comparison of scan data and simulation data according to claim 1, wherein the process step setting procedure comprises: And setting the process steps to be controlled according to the process key point control of the manufacturing and construction processes.
3. The segment deformation prediction method based on comparison of scan data and simulation data according to claim 1, wherein S2 specifically comprises: Performing coarse positioning transformation on the point cloud data and the patch model to obtain a rotation transformation matrix between the point cloud data and the patch model, and transforming the point cloud data according to the rotation transformation matrix to obtain coarse positioned point cloud data; And calculating the nearest point from each control point in the roughly positioned point cloud data, and obtaining the vector from the point to the control point to obtain the actually measured deformation.
4. A segment deformation prediction method based on scan data versus simulation data according to claim 3, wherein the coarse localization transformation employs ICP method.
5. A method of segment deformation prediction based on comparison of scan data with simulation data according to claim 3, wherein the point cloud data is obtained by scanning the bridge segment in the corresponding process step using a three-dimensional scanning technique, the three-dimensional scanning technique including at least one of laser scanning, oblique photography, and total station scanning.
6. The segment deformation prediction method based on comparison of scan data and simulation data according to claim 1, wherein the control points comprise key control points and auxiliary control points, the key control points comprise corner points of bridge segments, upper and lower edge points at intervals of one segment and the centroid of a slab, and the auxiliary control points comprise points generated according to uv values of the patches in a patch model or according to preset mesh intervals.
7. The segment deformation prediction method based on comparison of scan data and simulation data according to claim 6, wherein the deformation prediction model adopts a layered prediction structure and comprises two multi-layer perceptrons, wherein the multi-layer perceptrons F θ1 are used for extracting implicit features in a data set to obtain actual deformation of the auxiliary control points, and the multi-layer perceptrons F θ2 are used for fusing the implicit features with features of the data set to obtain actual deformation of the key control points.
8. The segment deformation prediction method based on comparison of scan data and simulation data according to claim 7, wherein S5 specifically comprises: After a control point is set for the bridge segment to be predicted, performing deformation simulation to obtain the simulation deformation of the control point; and inputting the simulation deformation into the deformation prediction model to obtain a deformation prediction result of the key control point.
9. The segment deformation prediction method based on comparison of scanning data and simulation data according to claim 8, wherein the deformation prediction result is used for indicating assembly of bridge segments in a hoisting deformation state, and process deformation comparison of the bridge segments after thermal process treatment, and orthopedic installation of the bridge segments.

Description

Segment deformation prediction method based on comparison of scanning data and simulation data Technical Field The invention relates to the technical field of bridge engineering, in particular to a segment deformation prediction method based on comparison of scanning data and simulation data. Background In the actual assembly and hoisting process of the large complex bridge, the requirements on high-precision matching and alignment of the sizes of the multiple sections are extremely high, but the cross section size is large, and after the thermal processing technology is carried out under different stress states, local and integral deformation exists, wherein part of the deformation can be recovered, and part of the deformation is plastic deformation. The technological process can be guided through simulation data, the deformation can be reduced, or the related deformation is utilized, so that the effect of smooth assembly and butt joint is achieved, and the butt joint construction difficulty is greatly reduced. However, the existing simulation input design is often too ideal and has large deviation from the actual deformation. Therefore, a prediction method capable of reflecting deformation of a segment member in each state relatively truly is required. Disclosure of Invention The invention aims to solve the problems of large deformation prediction deviation and the like in the existing bridge segment assembling process, and provides a segment deformation prediction method based on comparison of scanning data and simulation data. The invention provides a segment deformation prediction method based on comparison of scanning data and simulation data, which comprises the following steps: s1, setting control points based on a surface patch model of a bridge segment, and carrying out deformation analysis on the surface patch model according to set process steps to obtain simulation deformation of the control points in the corresponding process steps; s2, acquiring the actually measured deformation of the control point and the three-dimensional coordinate data of the control point based on the point cloud data of the bridge segment in the corresponding process step; s3, repeating the steps S1 to S2, and obtaining a plurality of simulation deformation amounts, actual measurement deformation amounts and three-dimensional coordinate data corresponding to the control points to generate a data set; S4, inputting the data set into the constructed neural network model for training to obtain a trained deformation prediction model; and S5, carrying out deformation prediction on the bridge segment to be predicted according to the deformation prediction model. According to a specific embodiment, in the deformation prediction method, the setting procedure of the process step includes: And setting the process steps to be controlled according to the process key point control of the manufacturing and construction processes. According to a specific embodiment, in the deformation prediction method, the S2 specifically includes: Performing coarse positioning transformation on the point cloud data and the patch model to obtain a rotation transformation matrix between the point cloud data and the patch model, and transforming the point cloud data according to the rotation transformation matrix to obtain coarse positioned point cloud data; And calculating the nearest point from each control point in the roughly positioned point cloud data, and obtaining the vector from the point to the control point to obtain the actually measured deformation. According to a specific embodiment, in the deformation prediction method, the coarse positioning transformation adopts an ICP method. According to a specific embodiment, in the deformation prediction method, the point cloud data is obtained by scanning the bridge segment in the corresponding process step by using a three-dimensional scanning technology, and the three-dimensional scanning technology includes at least one of laser scanning, oblique photography and total station scanning. According to a specific implementation mode, in the deformation prediction method, the control points comprise key control points and auxiliary control points, the key control points comprise corner points of bridge sections, upper and lower edge points at intervals and the mass center of a plate, and the auxiliary control points comprise points generated according to uv values of the plates in the plate model or according to preset mesh intervals. According to a specific embodiment, in the deformation prediction method, the deformation prediction model adopts a layered prediction structure and comprises two multi-layer perceptrons, wherein the multi-layer perceptrons F θ1 are used for extracting implicit features in a dataset to obtain the actual deformation of the auxiliary control point, and the multi-layer perceptrons F θ2 are used for fusing the implicit features with features of the dataset to obtain the a