Bridge finite element model construction method based on three-dimensional laser point cloud

Through high-precision three-dimensional laser scanning and automated processing algorithms, high-quality bridge geometric models are generated, which solves the problem of inefficient traditional construction monitoring and finite element modeling, and realizes rapid and timely evaluation and analysis of bridge structures.

CN118965887APending Publication Date: 2024-11-15SHANGHAI CIVIL ENG GRP CO LTD OF CREC +2

Patent Information

Application Number
CN202411029865.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

Traditional construction monitoring methods are difficult to comprehensively investigate the risk of bridge structures. Finite element modeling relies on manual operations and is time-consuming, making it difficult to meet the needs of rapid evaluation. Three-dimensional laser scanning technology provides high-precision point cloud data, but the data volume is huge, and traditional processing methods are difficult to process efficiently. .

Method used

High-precision 3D laser scanner is used to obtain bridge 3D point cloud data, remove noise and redundancy through automated algorithms, fill holes, repair defects, extract feature lines and key nodes, generate high-quality geometric models, and perform them in finite element analysis software. Meshing and attribute assignment enable rapid conversion from point cloud to finite element model.

Benefits of technology

It significantly improves modeling efficiency and geometric accuracy, realizes effective integration of field data and precise processing of complex nodes, provides strong support for the rapid evaluation of bridge structures, and ensures the accuracy and reliability of finite element analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965887A_ABST
    Figure CN118965887A_ABST
Patent Text Reader

Abstract

The invention discloses a rapid bridge finite element model construction method based on three-dimensional laser point cloud, and the method comprises the steps: carrying out the omnibearing scanning of a bridge through a high-precision three-dimensional laser scanner, and obtaining complete three-dimensional point cloud data; importing point cloud data obtained by multiple times of scanning into processing software, and carrying out data splicing, alignment and fusion; performing fine processing on the spliced point cloud model to ensure geometric accuracy; exporting the processed point cloud model in a geometric file format, importing the geometric model into finite element analysis software, and performing grid division; and finally, endowing the model with material attributes, boundary conditions and load conditions, and rapidly obtaining stress and deformation data of the bridge structure through static analysis. The method can be used for quickly constructing the finite element model of the construction bridge, the modeling efficiency and geometric accuracy are greatly improved, a solid data basis is provided for construction control of a bridge structure, and the safety and reliability of bridge engineering are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a finite element modeling method, in particular to a bridge finite element modeling method based on three-dimensional laser point cloud. Background Art

[0002] As an important infrastructure, the structural safety of bridges is of vital importance, and the safety of bridges during the construction process is of paramount importance. During the construction phase, the bridge structure is the most fragile and the stress is the most complex. Any minor mistakes or hidden dangers that are not discovered in time may lead to serious safety accidents, endanger the lives of construction workers, cause huge economic losses, and delay the progress of the project. Traditional construction monitoring methods are difficult to timely and comprehensively check structural risks, especially for the accurate estimation of structural stress changes. Finite element analysis, as a numerical analysis method, is widely used in bridge structure stress assessment. With the improvement of computing power, bridge finite element analysis has made great progress in dynamic analysis and nonlinear analysis of large-scale complex structures. However, finite element modeling in actual engineering usually relies on manual geometric modeling and meshing, which is cumbersome and time-consuming. For large-scale bridge structures, the modeling workload is huge and it is difficult to meet the needs of rapid evaluation. In addition, since the modeling method mainly relies on static design data such as design drawings, it is difficult to fully reflect the actual geometric shape and detail changes of the bridge, resulting in delayed updates of the finite element model and affecting the timeliness of the analysis results. 3D laser scanning technology can achieve real-time acquisition of on-site bridge morphology and generate digital models in the form of 3D point clouds. However, although this technology can provide high-precision point cloud data, the amount of data is huge, and traditional data processing methods are difficult to process efficiently, which limits its practical application. Therefore, it is urgent to develop a method for rapid modeling of finite element models of construction bridges, timely and accurately acquire large-scale bridge morphology, and automatically import it into finite element analysis software to quickly generate finite element models for efficient analysis. Summary of the invention

[0003] Purpose of the invention: The present invention aims to provide a method for constructing a bridge finite element model based on three-dimensional laser point cloud, which not only significantly improves the modeling efficiency and geometric accuracy, but also realizes the effective integration of field data and the precise processing of complex nodes, providing strong support for the rapid evaluation of bridge structures.

[0004] Technical solution: The method for constructing a bridge finite element model based on three-dimensional laser point cloud of the present invention comprises the following steps:

[0005] Use a high-precision 3D laser scanner to scan the bridge in all directions and at multiple angles to obtain complete 3D point cloud data of the bridge. During the scanning process, ensure that the data is fully covered to reduce data omissions and blind spots.

[0006] Based on the three-dimensional point cloud data of the bridge, the three-dimensional point cloud data of the bridge obtained by multiple scans is imported into the point cloud processing software for data splicing, alignment and fusion to generate a complete point cloud model of the bridge. The noise and redundant data are removed by the automated algorithm, the fine structure information is retained, and the accuracy and integrity of the point cloud model are improved.

[0007] Based on the bridge point cloud model, the spliced ​​point cloud model is refined, including filling holes, repairing defects, extracting feature lines and key nodes, etc. Advanced algorithms are used to ensure the geometric accuracy of the point cloud model and provide high-quality basic data for subsequent finite element modeling.

[0008] Export the processed point cloud model in a geometric file format. Use the conversion software to automatically identify the structural features in the point cloud and generate a geometric model that can be used for finite element analysis.

[0009] Import the generated geometric model into the finite element analysis software for meshing. Select the appropriate mesh type and density according to the analysis requirements to ensure the accuracy and calculation efficiency of the model.

[0010] Based on the finite element model, material properties, boundary conditions and load conditions are assigned to the model in the finite element analysis software to complete the construction of the finite element model. Static analysis is performed to quickly obtain key data such as stress and deformation of the bridge structure through an automated analysis process.

[0011] Furthermore, the process of scanning a bridge using a 3D laser scanner includes:

[0012] The optimal measurement point plan is obtained based on the point planning algorithm at the bridge site to ensure that the scanning equipment can cover every key part of the bridge. During the scanning process, the quality of the collected data is monitored in real time, and the point cloud data is analyzed in real time through the built-in quality assessment algorithm to detect and feedback missed areas and accuracy issues in the data.

[0013] Furthermore, the automatic algorithm for removing noise is a statistical filter algorithm, which includes calculating the distance mean and standard deviation of each point to its neighboring points, setting a certain multiple of the threshold, and removing outliers whose mean and standard deviation exceed the threshold. The formula is as follows:

[0014] Distance within the neighborhood: d ij =||p i -p j ||

[0015] Average distance within the neighborhood:

[0016] Standard deviation of distance within a neighborhood:

[0017] Outlier determination: μi >μ+ασorμ i <μ-ασ

[0018] Among them, d ij is the distance from the neighborhood point to the center point, p i 、p j is the point cloud data point, μ i For all points in the neighborhood to p i The average distance between points, k is the number of points in the specified neighborhood, σ i For all points in the neighborhood to p i The standard deviation of point distance, α is the threshold multiple.

[0019] Furthermore, the automated algorithm for removing redundant data is a random sampling consistency algorithm, which includes randomly selecting a subset of the point cloud to fit the model, evaluating the fitness of the model with all points, calculating the number of inliers, iterating multiple times, and selecting the model with the largest number of inliers as the final model. The formula is as follows:

[0020] Model fitting equation: az+by+cz+d=0

[0021] Error calculation:

[0022] Interior point determination: e i ≤∈

[0023] Iteration number estimation:

[0024] Among them, (a, b, c) is the plane normal vector, d is the plane distance constant, e i is the point cloud data p i The distance error to the fitted model, (x i ,y i ,z i ) is p i The point coordinates, ∈ is the error threshold, N is the number of iterations, P is the success probability, p is the proportion of inliers, and m is the minimum number of points required for model fitting.

[0025] Furthermore, the hole filling includes using the neighborhood information of the point cloud data to automatically fill the holes in the point cloud model through a multi-scale neighborhood interpolation algorithm. The defect repair includes repairing the details in the point cloud data using surface fitting technology and local reconstruction algorithm. The feature line and key node extraction includes extracting the key feature points and structural lines of the bridge from the point cloud data through morphological analysis and edge detection algorithm.

[0026] Furthermore, the point cloud model is converted into a geometric model by using B-spline curve and NURBS parametric modeling technology.

[0027] Furthermore, the meshing is achieved through an adaptive meshing algorithm.

[0028] Furthermore, the meshed geometric model is assigned steel or concrete material properties and unit types according to the component type, constraints are imposed on the corresponding nodes according to the actual boundary conditions, self-weight and vehicle loads are added to the model, and static analysis is performed using finite element analysis software. The stress and deformation data of key positions are automatically extracted for subsequent mechanical performance evaluation.

[0029] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: This method not only significantly improves the modeling efficiency and geometric accuracy, but also realizes the effective integration of field data and the precise processing of complex nodes, providing strong support for the rapid evaluation of bridge structures. First, the present invention introduces high-precision three-dimensional laser scanning technology and automated processing algorithms to achieve efficient processing of point cloud data and rapid generation of geometric models. The automated algorithm removes noise and redundant data, fills holes, repairs defects, and extracts key structural features, significantly improving the accuracy and integrity of the point cloud model. It not only reduces the workload of manual operations, but also ensures the generation of high-quality geometric models, providing a solid foundation for subsequent finite element modeling; secondly, the present invention quickly constructs a refined bridge finite element model through the seamless integration of point cloud models and finite element analysis software, solving the problems of difficult and inefficient model conversion. Through conversion, the structural features in the point cloud are accurately identified and a geometric model is generated. Then, meshing, attribute assignment and boundary condition setting are performed in the finite element analysis software, realizing efficient conversion from point cloud data to finite element model, greatly improving modeling efficiency and geometric accuracy, and ensuring the accuracy and reliability of finite element analysis. Finally, the method of the present invention solves the problems of difficult rapid processing of construction site data and inaccurate processing of complex nodes through efficient point cloud data processing, refined geometric model generation and seamless finite element analysis integration. Through the automated analysis process, key data such as stress and deformation of the bridge structure are quickly obtained, realizing rapid and timely evaluation of the bridge structure during the construction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the method for rapidly constructing a bridge finite element model based on three-dimensional laser point cloud according to the present invention;

[0031] Figure 2 The generated three-dimensional point cloud image of the bridge;

[0032] Figure 3 Convert the bridge geometry model into a finite element model diagram;

[0033] Figure 4 This is the stress-displacement distribution diagram of the bridge main beam. DETAILED DESCRIPTION

[0034] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0035] The method for rapid construction of bridge finite element model based on three-dimensional laser point cloud described in the present invention is shown in the flowchart. Figure 1 , including the following steps:

[0036] (1) Select a high-precision 3D laser scanner suitable for bridge structures and set up multiple scanning points at different locations on the bridge to cover the entire bridge structure. Each scanning point should be reasonably arranged according to the geometric shape and environmental conditions of the bridge to ensure that the scanning angle and range can fully cover all parts of the bridge. In particular, for the key nodes, connection parts and complex structural parts of the bridge, detailed scanning is required to capture their detailed geometric information.

[0037] (2) Import the acquired point cloud data from multiple scans into the point cloud processing software and stitch the data from multiple scans together. Use the automatic registration function of the software to accurately stitch together data from different scanning perspectives through feature point matching and overlapping area alignment. Use the global registration algorithm to globally align the initially stitched point cloud data to generate a complete bridge point cloud model. Use the automated algorithm to remove noise and redundant data, retain fine structure information, and improve the accuracy and integrity of the point cloud model.

[0038] (3) Based on the bridge point cloud model obtained in step (2), the spliced ​​point cloud model is refined, including filling holes, repairing defects, extracting feature lines and key nodes, etc. Advanced algorithms are used to ensure the geometric accuracy of the point cloud model and provide high-quality basic data for subsequent finite element modeling.

[0039] (4) In the point cloud processing software, set the export parameters, select the appropriate geometry file format (such as STL, OBJ, STEP, etc.), convert and save the processed point cloud model to the selected geometry file format, and perform geometric consistency verification on the generated geometry model to ensure that the model is consistent with the original point cloud data. If not, the point cloud model needs to be further refined through step 3. Automatically identify key structural features in the geometry model, including beams, columns, bridge decks, connection nodes, etc.

[0040] (5) Based on the geometric model generated in step (4), the model is imported into finite element analysis software for meshing. According to the analysis requirements, a suitable mesh type and density are selected to ensure the accuracy and computational efficiency of the model. The mesh types include line unit mesh, surface unit mesh and volume unit mesh.

[0041] (6) For the finite element model of step (5), select the appropriate material model according to the actual material of the bridge, input the specific parameters of the material, assign them to each part of the model, and set constraints at the corresponding positions. According to the load conditions of the bridge in actual use, select the appropriate load type, define the load size and action location. Perform static analysis and quickly obtain key data such as stress and deformation of the bridge structure through an automated analysis process.

[0042] Taking a six-span simply supported concrete beam bridge as an example, the specific process of the method for rapid modeling of the bridge finite element model based on three-dimensional laser point cloud is given:

[0043] (1) 3D laser scanning

[0044] The Leica RTC360 high-precision 3D laser scanner was selected, with an accuracy within ±1 mm. According to the length and complexity of the bridge, a total of 24 scanning points were set. Four scanning points were set for each span of the bridge, and a total of 24 scanning points for the six spans of the bridge to ensure that the entire area of ​​the bridge was covered. The scanning resolution was set to 10 million points per 1,000 square meters, and the scanning speed was set to 1 million points per second to ensure high-precision data collection. The coverage of the scanner was adjusted so that the scanning radius of each scanning point was 100 meters to ensure that the overlapping area of ​​adjacent scanning points was large enough for subsequent data splicing. Scanning was performed at 24 scanning points in sequence to ensure that the data coverage of each scanning point was comprehensive and to reduce data omissions and blind spots. Detailed scanning was performed at multiple angles at the connection between the pier and the beam to capture detailed information of the connection. A comprehensive scan of the beam bottom and the support system was performed to ensure that the geometric information of these complex structures was fully recorded.

[0045] (2) Bridge point cloud generation

[0046] The point cloud data obtained from the 24 scanning points were imported into the point cloud processing software Leica Cyclone to ensure that the imported data format was compatible. The data volume for each scan was about 500GB. The automatic registration function of the software was used to stitch the data from the 24 scanning angles together by matching feature points and aligning the overlapping areas (about 20% of the area of ​​each scan). Figure 2. The preliminary spliced ​​point cloud data are globally aligned using the global registration algorithm (ICP). The algorithm minimizes the overall error of each scanned point cloud data through iterative calculation. The number of iterations of the ICP algorithm is set to 50 times, and the convergence criterion is 0.001 meters. The statistical filter algorithm is used to remove noise points in the point cloud data. The number of neighborhood points of each point is set to 50, and the standard deviation multiple threshold is set to 2.0, and noise points whose distance from the mean exceeds 2 times the standard deviation are removed. The random sampling consensus algorithm (RANSAC) is used to identify and remove redundant points in the point cloud data. The number of iterations of the RANSAC algorithm is set to 1000 times, and the inlier threshold is set to 0.05 meters, that is, points with a fitting model error within 0.05 meters are considered to be inliers.

[0047] (3) Fine-tuning the point cloud model

[0048] Identify the hole areas in the point cloud model. For each hole point, define its neighborhood as all points within a radius of 0.1 meters. In each neighborhood, based on the geometric information of the adjacent points, generate new filling points through multi-scale interpolation calculation to ensure the geometric consistency and continuity of the filled points with the surrounding points. For larger holes, fill them step by step in multiple times, using different scales each time, and increase the number and density of filling points layer by layer, and finally generate a smooth and seamless filling result. Select control points in the defective area and use surface fitting technology to repair the defective area. Use morphological operations such as dilation, erosion, opening and closing operations to highlight the structural features in the point cloud. Apply edge detection algorithm to identify the edges in the point cloud model and extract the characteristic lines of the bridge. Use key point detection algorithm to extract key points in the geometric model. Mark the extracted characteristic lines and key nodes to generate feature point cloud data.

[0049] (4) Geometry model conversion

[0050] In the point cloud processing software, the export resolution is set to 0.01 meters to ensure the balance between the model's precision and the amount of data, and the unit is set to meters to ensure consistency with the actual size. After setting the parameters, the processed point cloud model is converted and saved as a geometric file in OBJ format. The point cloud data is parametrically modeled using B-spline curve technology, and 20 control points per meter are selected to fit the point cloud data. The generated geometric model is verified for geometric consistency using 3D visualization software, the error between the geometric model and the point cloud data is compared, and a comprehensive visual inspection of the geometric model is performed to ensure the integrity of the model and the absence of obvious defects. The components at each position of the geometric model are classified and marked as main beams, transverse connections, and supports.

[0051] (5) Unit grid division

[0052] The finite element analysis software ANSYS is used for meshing. The previously generated geometric model file is imported. In order to simplify the calculation, only one span is selected for meshing. Figure 3 . The initial mesh size is set to 0.1 meters to ensure the overall coverage of the model. Locally refined meshes are set in stress concentration areas (such as supports and connection nodes) and complex geometric areas (such as piers and bridge deck junctions), and the mesh size is adjusted to 0.02 meters. In areas with relatively uniform stress (such as the middle part of the bridge deck), a coarser mesh is used, and the mesh size is adjusted to 0.2 meters. Line element mesh is used for the beam part of the bridge, and the line element length is set to 0.1 meters. Use the quality check function of the software to evaluate the quality of the mesh elements and ensure that the quality of all elements is within an acceptable range.

[0053] (6) Finite element analysis

[0054] According to the actual bridge data, in the finite element analysis software, the concrete material properties are assigned to the main beam, and the constraints are applied to the corresponding nodes and boundaries of the model through the graphical interface operation. The self-weight load is calculated according to the volume of the main beam and applied to each component. The vehicle load adopts the highway-I level load based on the actual bridge data. In the finite element analysis software, set the solution parameters, such as the number of iterations (set to 100 times) and the convergence criterion (set to 0.001). Through the post-processing module of the finite element analysis software, the calculation results are visualized, and internal force diagrams, deformation diagrams, etc. are generated to intuitively display the mechanical properties of the bridge under different working conditions, and extract key data in the calculation results (such as maximum internal force, maximum deformation, etc.) for reference. Figure 4 , based on the analysis results, judge the safety and reliability of the bridge structure.

Claims

1. A bridge finite element modeling method based on three-dimensional laser point cloud, It is characterized in that The steps include: Use a 3D laser scanner to scan the bridge and obtain complete 3D point cloud data of the bridge; Import the 3D point cloud data of the bridge obtained by multiple scans into the point cloud processing software, perform data splicing, alignment and fusion, generate a complete point cloud model of the bridge, and remove noise and redundant data through automated algorithms; Based on the bridge point cloud model, the spliced ​​point cloud model is refined, including filling holes, repairing defects, and extracting feature lines and key nodes; Export the processed point cloud model in a geometric file format, use conversion software to automatically identify the structural features in the point cloud, and generate a geometric model that can be used for finite element analysis; Import the generated geometric model into the finite element analysis software for meshing; Based on the finite element model, material properties, boundary conditions and load conditions are assigned to the model in the finite element analysis software, the construction of the finite element model is completed, static analysis is performed, and relevant data of the bridge structure is obtained through an automated analysis process.

2. The bridge finite element modeling method based on three-dimensional laser point cloud according to claim 1, It is characterized in that The process of scanning a bridge using a 3D laser scanner includes: The optimal measurement point plan is obtained based on the point planning algorithm at the bridge site to ensure that the scanning equipment can cover every key part of the bridge. During the scanning process, the quality of the collected data is monitored in real time, and the point cloud data is analyzed in real time through the built-in quality assessment algorithm to detect and feedback missed areas and accuracy issues in the data.

3. The bridge finite element modeling method based on three-dimensional laser point cloud according to claim 1, It is characterized in that The automatic algorithm for removing noise is a statistical filter algorithm, which includes calculating the distance mean and standard deviation of each point to its neighboring points, setting a threshold, and removing outliers whose mean and standard deviation exceed the threshold. The formula is as follows: Distance within the neighborhood: d ij =||p i -p j || Average distance within the neighborhood: Standard deviation of distance within a neighborhood: Outlier determination: μ i > μ + ασ or μ i < μ - ασ Among them, d ij is the distance from the neighborhood point to the center point, p i 、p j is the point cloud data point, μ i For all points in the neighborhood to p i The average distance between points, k is the number of points in the specified neighborhood, σ i For all points in the neighborhood to p i The standard deviation of point distance, α is the threshold multiple.

4. The bridge finite element modeling method based on three-dimensional laser point cloud according to claim 1, It is characterized in that The automated algorithm for removing redundant data is a random sampling consensus algorithm, which involves randomly selecting a subset of the point cloud to fit the model, evaluating the fitness of the model with all points, calculating the number of inliers, iterating multiple times, and selecting the model with the largest number of inliers as the final model. The formula is as follows: Model fitting equation: az+by+cz+d=0 Error calculation: Interior point determination: e i ≤∈ Iteration number estimation: Among them, (a, b, c) is the plane normal vector, d is the plane distance constant, e i is the point cloud data p i The distance error to the fitted model, (x i ,y i ,z i ) is p i The point coordinates, ∈ is the error threshold, N is the number of iterations, P is the success probability, p is the proportion of inliers, and m is the minimum number of points required for model fitting.

5. The method for constructing a bridge finite element model based on three-dimensional laser point cloud according to claim 1, It is characterized in that The hole filling includes using the neighborhood information of the point cloud data to automatically fill the holes in the point cloud model through a multi-scale neighborhood interpolation algorithm.

6. The bridge finite element modeling method based on three-dimensional laser point cloud according to claim 1, It is characterized in that The defect repairing includes repairing the details in the point cloud data by using surface fitting technology and local reconstruction algorithm.

7. The method for constructing a bridge finite element model based on three-dimensional laser point cloud according to claim 1, It is characterized in that The extracting of characteristic lines and key nodes includes extracting key characteristic points and structural lines of the bridge from point cloud data through morphological analysis and edge detection algorithms.

8. The method for constructing a bridge finite element model based on three-dimensional laser point cloud according to claim 1, It is characterized in that The conversion of point cloud model into geometric model is realized through B-spline curve and NURBS parametric modeling technology.

9. The method for constructing a bridge finite element model based on three-dimensional laser point cloud according to claim 1, It is characterized in that The grid division is achieved by an adaptive grid division algorithm.

10. The method for constructing a bridge finite element model based on three-dimensional laser point cloud according to claim 1, It is characterized in that The meshed geometric model is assigned steel or concrete material properties and unit types according to the component type. Constraints are applied to the corresponding nodes based on the actual boundary conditions. Self-weight and vehicle loads are added to the model. Static analysis is performed using finite element analysis software, and stress and deformation data at key locations are automatically extracted for subsequent mechanical performance evaluation.

Citation Information

Patent Citations

  • Complex building three-dimensional modeling method based on point cloud data

    CN106600690A

  • Three-dimensional entity model reconstruction method based on dense point cloud data

    CN111932671A

  • Pipeline maintenance method based on laser scanning and BIM technology

    CN113946915A

  • Reverse engineering scanner result processing method and device based on DBSCAN

    CN116310455A

  • Building structure finite element intelligent reverse modeling and analysis system based on three-dimensional computer vision

    CN116611280A

Cited By

  • Bridge structure modeling method based on three-dimensional point cloud processing

    CN119206102A

  • Structural health monitoring system and method for urban comprehensive pipe gallery

    CN119824965A

  • Rapid three-dimensional modeling method and system for steam engine room pipe network of thermal power plant

    CN120070751A

  • Aviation metal braided conductor point cloud data entity modeling method based on cellular corrosion

    CN120470827A

  • Bridge deformation detection method based on three-dimensional laser scanning and multi-scale algorithm

    CN120807485A