Bridge model intelligent construction method and system based on BIM

Abstract

The embodiment of the invention provides a BIM-based bridge model intelligent construction method and system, and belongs to the field of BIM-based bridge model intelligent construction. The method comprises the steps that bridge model construction related data are collected and preprocessed, and a component library is established according to the preprocessed bridge model construction related data; defining geometric linkage rules and topological connection logic of components in the component library, and generating an initial BIM bridge model through a generative adversarial network; performing lightweight finite element calculation on the initial BIM bridge model to obtain an optimized BIM bridge model; and performing standard conflict detection and conflict parameter adjustment on the optimized BIM bridge model to obtain a final bridge model. According to the method, the parameter relation matrix and the dynamic linkage mechanism are constructed, the generative adversarial network is introduced to generate the complex bridge type, and the graph convolutional network is introduced into the generative adversarial network, so that reasonable bridge type topology and dynamic parameter linkage are guaranteed.

Description

Technical Field

[0001] The present invention relates to the intelligent construction of a bridge model based on BIM, and in particular to a method and system for the intelligent construction of a bridge model based on BIM. Background Art

[0002] In recent years, information technology and intelligent technologies have rapidly developed in the field of bridge engineering. Building Information Modeling (BIM) has become a crucial tool for bridge lifecycle management. Traditional BIM modeling methods primarily rely on manual operations, constructing geometric models from CAD drawings. This approach is inefficient and prone to errors. With the widespread adoption of technologies such as 3D laser scanning and drone photogrammetry, reverse modeling methods based on point cloud data have significantly improved the accuracy of bridge geometric information acquisition. However, multi-source data fusion still faces challenges. For example, fusion methods based on feature point matching are susceptible to noise in complex scenarios, while the iterative closest point (ICP) algorithm is sensitive to initial registration and can lead to local optimality. In terms of parametric modeling, mainstream BIM platforms (such as Revit and Tekla) enable dynamic adjustment of component morphology through family parameterization. However, existing parameter relationship representations are mostly based on linear logic, making it difficult to support the automated updating of complex, nonlinearly dependent parameters. Furthermore, while the integration of finite element analysis and BIM has matured, traditional finite element models consume large computational resources due to their fixed mesh density, making them particularly inefficient when dealing with long-span bridges. Summary of the Invention

[0003] The purpose of the embodiments of the present invention is to provide a BIM-based bridge model intelligent construction method and system for efficient collaboration and intelligent optimization of the entire bridge design process.

[0004] To achieve the above-mentioned objectives, an embodiment of the present invention provides a BIM-based intelligent bridge model construction method, comprising: collecting and preprocessing bridge model construction-related data, establishing a component library based on the preprocessed bridge model construction-related data; defining geometric linkage rules and topological connection logic of components in the component library, and generating an initial BIM bridge model through a generative adversarial network; performing lightweight finite element calculation on the initial BIM bridge model to obtain an optimized BIM bridge model; and performing specification conflict detection and conflict parameter adjustment on the optimized BIM bridge model to obtain a final bridge model.

[0005] Optionally, the bridge model construction related data includes: spatial coordinate data, design input data, historical bridge BIM model data and construction method case library data.

[0006] Optionally, defining the geometric linkage rules and topological connection logic of the components in the component library includes: extracting bridge parameters from historical bridge BIM model data and construction method case library data, and dividing the bridge parameters into independent parameters and dependent parameters; establishing a parameter relationship matrix based on the independent parameters and the dependent parameters; performing mathematical relationship conversion according to the parameter relationship matrix to obtain mathematical relationship expressions of each bridge component; embedding the mathematical relationship expressions of each bridge component into the BIM platform to obtain a dynamically linked parametric component family.

[0007] Optionally, the definition of the geometric linkage rules and topological connection logic of the components in the component library also includes: classifying the connection types of the components, and prioritizing the classified connection types, wherein the priority ranking includes design priority, calculation priority, conflict resolution priority, and maintenance priority; modeling constraints according to the priority ranking to obtain component connection constraints; embedding the component connection constraints into the BIM platform, and performing dynamic connection verification.

[0008] Optionally, generating an initial BIM bridge model through a generative adversarial network includes: constructing a training data set for the generative adversarial network; constructing a generator and a discriminator for the generative adversarial network; designing a loss function and formulating a training strategy, and according to the formulated training strategy, using the training data set to train the generator and the discriminator in stages; mapping and converting the bridge design parameters output by the generator, using BIM to construct a bridge model, and obtaining an initial BIM bridge model.

[0009] Optionally, constructing the generator and discriminator of the generative adversarial network includes: introducing a graph convolutional network into both the generator and the discriminator, and using the graph convolutional network in the generator to encode the input graph structure data to obtain an embedded representation of parameterized components and the connection types of the components; regenerating the geometric parameters and topological structure of the bridge type based on the embedded representation; the discriminator processes the graph structure data of the bridge type through the graph convolutional network, and learns the feature representation of the parameterized components and the connection types of the components.

[0010] Optionally, the lightweight finite element calculation is performed on the initial BIM bridge model to obtain an optimized BIM bridge model, including: geometric simplification and feature extraction of the initial BIM bridge model; adaptive mesh optimization of the simplified initial BIM bridge model to obtain a lightweight BIM bridge model; solver acceleration of the lightweight BIM bridge model and extraction of mechanical indicators; multi-objective optimization of the lightweight BIM bridge model after solver acceleration to obtain an optimized BIM bridge model.

[0011] Optionally, the multi-objective optimization includes: using the non-dominated sorting genetic algorithm II generation to randomly generate an initial population, where each individual represents a parameterized combination of a bridge design scheme; calculating the multi-objective function value of each individual and performing non-dominated sorting; calculating the individual crowding distance in each non-dominated sorting level, using the tournament selection method to select individuals from the current population as parent individuals; using a simulated binary crossover operator to perform a crossover operation on the selected parent individuals; using a polynomial mutation operator to perform a mutation operation on the offspring individuals, and checking whether the new individuals meet the constraints.

[0012] Optionally, after performing code conflict detection and conflict parameter adjustment on the optimized BIM bridge model, a final bridge model is obtained, including: the conflict detection includes geometric conflict detection and code conflict detection; the adjustment includes geometric adjustment, parameter correction and iterative verification.

[0013] On the other hand, the present invention provides a BIM-based bridge model intelligent construction system for implementing a BIM-based bridge model intelligent construction method. The system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. The processor executes the computer program to implement the BIM-based bridge model intelligent construction method.

[0014] The above technical solution improves the point cloud preprocessing algorithm, constructs a parameter relationship matrix and dynamic linkage mechanism, introduces a generative adversarial network to generate complex bridge types, and introduces a graph convolutional network into the generative adversarial network to ensure reasonable bridge topology, dynamic parameter linkage, and automatic compliance with regulations. It combines adaptive meshes with CUDA-accelerated lightweight finite element analysis to improve computing efficiency, shorten calculation time, and a multi-objective optimization-driven specification conflict adjustment strategy, thus achieving efficient collaboration and intelligent optimization of the entire bridge design process.

[0015] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation of the embodiments of the present invention. In the accompanying drawings: Figure 1 It is a flowchart of intelligent construction of bridge model based on BIM.

[0017] Figure 2 This is the optimization flow chart of the second generation of non-dominated sorting genetic algorithm. DETAILED DESCRIPTION

[0018] The following is combined with Figure 1 -Attached Figure 2 The specific implementation of the embodiment of the present invention is described in detail. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application are in compliance with the relevant provisions of national laws and regulations. In the embodiments of this application, certain software, components, models, and other existing solutions in the industry may be mentioned. These should be considered as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of this application, but it does not mean that the applicant has or will necessarily use such solutions.

[0020] In the process of realizing the present invention, the inventors of the present application discovered that the existing bridge BIM modeling technology has shortcomings such as weak parametric logic dynamics, low level of intelligent generation of complex bridge types, and low efficiency of finite element calculations.

[0021] Example 1 Reference Figure 1-Figure 2 , which is the first embodiment of the present invention, provides a method for intelligently constructing a bridge model based on BIM, comprising: S100: collecting and preprocessing bridge model construction related data, and establishing a component library based on the preprocessed bridge model construction related data.

[0022] Specifically, different devices such as drones, ground-based 3D laser scanners, total stations, and GPS are used to collect spatial coordinate data, which are then pre-processed and fused. The pre-processed and fused data are geographic-related data, and the fusion method can use methods such as feature point matching and ICP algorithm (Iterative Closest Point) to obtain unified spatial coordinate point cloud data.

[0023] Preferably, the spatial coordinate point cloud data is denoised, and a statistical filtering algorithm is used to calculate the neighborhood point cloud density of each point in the spatial coordinate data and remove density anomalies; a radius filtering algorithm is used to remove isolated points with insufficient number of points in the local area.

[0024] Preferably, the denoised spatial coordinate point cloud data is subjected to point cloud downsampling to reduce the number of point clouds and retain key geometric features.

[0025] Furthermore, design input data should be collected, including CAD (Computer-Aided Design) drawings related to bridge model construction. The collected CAD drawings should include detailed geometric information such as the bridge's plan layout, elevation, and cross-section, such as the bridge's span, deck width, and pier location. Load demand data should be obtained, including dead loads (such as the bridge's own weight), live loads (such as vehicle loads and crowd loads), wind loads, and seismic loads. Specification documents should be collected, which should include various standards and requirements for bridge design, including structural dimensions, material properties, safety factors, and other aspects.

[0026] Furthermore, by organizing the BIM model data of historical bridges, we can extract existing bridge component parameters, structural systems, design experience and other information, provide reference and reference for the design of new bridges, and improve design efficiency and quality.

[0027] Furthermore, collecting construction method case database data and understanding the characteristics, difficulties and solutions of different bridge types during the construction process will help to fully consider the construction feasibility in the design stage, optimize the bridge design plan, and reduce changes and risks during the construction process.

[0028] Preferably, after all data are collected, they are pre-processed again, and the pre-processing uses the data after the initial pre-processing and other data except the geographical data. A bridge design rule library is established based on the historical bridge BIM model data and the construction method case library data.

[0029] Furthermore, the components are classified according to their functions and forms. The classification of common bridge model components is shown in Table 1.

[0030]

[0031] It should be noted that the classification of bridge model components is not limited to the contents in Table 1.

[0032] Furthermore, the shape of each bridge model component is controlled by key dimensional parameters (length, width, height, thickness, etc.). For example, the cross-section of a box girder is precisely defined by parameters such as top plate thickness, web spacing, and beam height. Changing these dimensional parameters allows the box girder's geometry to be quickly adjusted to suit different design requirements. During modeling, parametric tools such as Revit's family editing function or Tekla's macro function are used to set component geometric parameters as variable variables, enabling dynamic adjustment of the component by adjusting these geometric parameters.

[0033] Furthermore, by binding component material properties, the system automatically associates material characteristics (such as concrete strength and steel grade) with the component, and automatically calculates physical properties (mass and moment of inertia) based on the material's mechanical properties. For example, for a concrete main beam, the system automatically calculates the mass per unit volume and the moment of inertia of the cross section based on the concrete strength grade, providing basic data for subsequent structural analysis and design.

[0034] Preferably, a material database is established and characteristic parameters of common materials are entered therein to facilitate calling and binding during the modeling process while ensuring the accuracy and consistency of material properties.

[0035] Furthermore, the pre-processed spatial coordinate point cloud data is unified with the coordinates of the CAD drawing, and the ICP algorithm is used to optimize the alignment accuracy (position accuracy, orientation accuracy, and scale accuracy) between the point cloud and CAD.

[0036] It's important to note that component properties must strictly adhere to the IFC (Industry Foundation Classes) standard. After verifying that component properties adhere to the IFC standard, the component library is established. During the component library establishment process, component categories, parameters, attributes, and other information are defined and stored according to the IFC standard. This ensures accurate information transfer when exchanging and sharing data between different software, avoiding data loss and format incompatibilities.

[0037] Optimally, component modification logs (e.g., "box girder web thickness changed from 0.6m to 0.8m") should be recorded, detailing the time, content, and reason for each modification, allowing for tracing component evolution. Historical version rollback is supported, preserving optimization process data. This facilitates performance comparisons between component versions or reverting to previous versions. Analyzing historical version data also provides a reference for subsequent design optimization.

[0038] Optimally, multi-device fusion sampling ensures the accuracy and reliability of spatial coordinate point cloud data, providing high-quality foundational data for subsequent modeling. Component classification based on historical BIM models, construction method case libraries, and CAD drawings, combined with parametric modeling and material property binding, enables rapid component reuse and dynamic adjustment, improving design efficiency. Component properties adhere to the IFC standard, supporting cross-platform data exchange. Modification logs and version backtracking facilitate design optimization iterations and accountability tracking.

[0039] S200: defining geometric linkage rules and topological connection logic of components in the component library, and generating an initial BIM bridge model through a generative adversarial network.

[0040] Furthermore, bridge parameters are extracted from historical bridge BIM model data and construction method case library data, and the bridge parameters are divided into independent parameters and dependent parameters; a parameter relationship matrix is ​​established based on the independent parameters and the dependent parameters; mathematical relationship conversion is performed according to the parameter relationship matrix to obtain mathematical relationship expressions of various bridge components; the mathematical relationship expressions of various bridge components are embedded in the BIM platform to obtain a dynamically linked parametric component family.

[0041] Specifically, geometric parameters such as cross-sectional dimensions, length, and height are extracted from historical bridge BIM models; material parameters such as elastic modulus and yield strength are extracted; construction parameters such as construction sequence and prestress value are extracted from the construction method case library data; construction process parameters such as casting segments and assembly sequence are extracted.

[0042] Furthermore, independent parameters are values ​​that do not depend on other parameters, such as the total length of the bridge, the height of the main beam section, and the material type; dependent parameters are values ​​that depend on independent parameters or other dependent parameters, such as the pier spacing, which depends on the total length and number of spans of the bridge; and the number of prestressed tendons, which depends on the cross-sectional size and force requirements.

[0043] Preferably, the upstream independent parameters and intermediate dependent parameters of each dependent parameter are analyzed to ensure that the relationship is complete and the logic is clear.

[0044] Next, a parameter relationship matrix is ​​created, where rows represent independent parameters and columns represent dependent parameters. The mathematical relationships between the parameters are entered into the matrix to create a parameter relationship matrix. A comprehensive logical review is conducted after the matrix is ​​obtained to ensure that all relevant parameters and their dependencies are correctly included.

[0045] Furthermore, the mathematical expressions in the parameter relationship matrix are converted into mathematical expressions that can be embedded in BIM, and all mathematical expressions are unified into a format suitable for BIM parametric modeling. The converted mathematical expressions are verified to ensure that they are consistent with the actual bridge design logic. If there are any deviations, the mathematical expressions are adjusted.

[0046] Furthermore, a parametric component family is created in the BIM platform, independent parameters are used as parametric component family parameters, mathematical expressions are embedded in the parametric component family parameters, and independent parameters are editable parameters; dependent parameters are automatically calculated and assigned by mathematical expressions and are not exposed as editable parameters to avoid circular dependencies.

[0047] Furthermore, by modifying independent parameters, it is verified whether dependent parameters are automatically updated, and multi-level dependency tests are performed. For example, changes in the main beam cross-sectional dimensions lead to changes in deadweight, resulting in a dynamically linked parametric component family that supports rapid design changes and optimization. The generated parametric component family is saved for subsequent use.

[0048] Furthermore, the connection types of the components are classified and prioritized, wherein the priority ranking includes design priority, calculation priority, conflict resolution priority and maintenance priority; constraint condition modeling is performed according to the priority ranking to obtain component connection constraint conditions; the component connection constraint conditions are embedded in the BIM platform, and dynamic connection verification is performed.

[0049] Specifically, the component connection types are classified (e.g., welding, bolted connection, hinged connection), and the connection types are prioritized according to design priority (e.g., structural safety), calculation priority (e.g., force analysis), conflict resolution priority (e.g., collision detection), and maintenance priority (e.g., maintainability).

[0050] Furthermore, based on the priority of the connection types, constraints are established for each connection type to obtain the component connection constraints. For example, welded connections require component alignment and no gaps, while bolted connections require matching apertures. Using BIM's constraint functions (such as Revit's alignment constraints and isometric constraints), the connection type constraints are embedded in the model, and the validity of the component connection constraints is tested in BIM. After the test is completed, connections that meet the component connection constraints are retained, and connections that do not meet the component connection constraints are repaired, such as parameter adjustment, topological relationship correction, material property rebinding, and construction process optimization, to obtain a BIM model with dynamic connection verification.

[0051] Preferably, if the bridge type belongs to a standardized bridge in the template library, such as a simply supported bridge, a T-shaped rigid frame bridge, etc., based on the input parameters (span combination, load level), the bridge template in the bridge design rule library is called, the pier position is automatically matched according to the spatial coordinate point cloud data, and a parametric bridge type scheme is output. If the bridge type does not belong to a standardized bridge in the template library, the bridge type is determined to be a complex bridge type.

[0052] Furthermore, for complex bridge types, such as special-shaped cable-stayed bridges and spatial curved bridge towers, the corresponding geometric parameter data, topological relationship data and mechanical property data are extracted from the pre-processed historical bridge BIM model library data and construction method case library data as training data.

[0053] Furthermore, a training data set for the generative adversarial network is constructed; a generator and a discriminator of the generative adversarial network are constructed; a loss function is designed and a training strategy is formulated, and according to the formulated training strategy, the generator and the discriminator are trained in stages using the training data set; the bridge design parameters output by the generator are mapped and converted, and a bridge model is constructed using BIM to obtain an initial BIM bridge model.

[0054] Furthermore, the construction of the generator and discriminator of the generative adversarial network includes: introducing a graph convolutional network into both the generator and the discriminator, and using the graph convolutional network in the generator to encode the input graph structure data to obtain an embedded representation of parameterized components and the connection types of components; regenerating the geometric parameters and topological structure of the bridge type based on the embedded representation; the discriminator processes the graph structure data of the bridge type through the graph convolutional network, and learns the feature representation of parameterized components and the connection types of components.

[0055] Specifically, a generative adversarial network (GAN) is used to design a generator and discriminator. A graph convolutional network (GCN) is incorporated into the generator. Training data is fed into the generator and organized into a topological graph structure. Nodes in this topological graph represent individual bridge components (such as main beams, towers, and cables), while edges represent the connections between components. The feature vectors of nodes contain geometric parameters and mechanical properties, while the feature vectors of edges contain the type and properties of the connection. By incorporating a GCN into this graph structured data, the GCN learns feature representations for both nodes and edges, capturing the complex relationships between bridge components.

[0056] Furthermore, in the generator, the input graph structure data is first encoded through a graph convolutional network to obtain embedded representations of nodes and edges. The embedded representations of nodes and edges are then used to generate the geometric parameters and topological structure of the new bridge type. The output of the generator is a bridge type scheme, which includes geometric parameters and topological relationships such as the cross-sectional shape of the main beam, the shape of the tower, and the arrangement of the cables. The bridge type scheme generated by the generator and the collected real bridge type data are input into the discriminator to evaluate the similarity between the bridge type scheme generated by the generator and the real bridge type data, and to judge the rationality of the bridge type scheme. The discriminator also introduces a graph convolutional network to process the graph structure data of the bridge type, learn the feature representations of nodes and edges, and by comparing the graph structure features of the generated scheme and the real data, the discriminator outputs the rationality probability of the generated bridge type scheme.

[0057] Preferably, a real-time specification checking module is embedded in the discriminator to ensure that the generated bridge type scheme meets the design standards.

[0058] Furthermore, if the discriminator considers the solution unreasonable and the rationality probability of the generated bridge solution is less than the set rationality threshold, the generator will update the network parameters of the generator and discriminator through the back-propagation algorithm to optimize the generation effect of the generator. After multiple iterative optimizations, the final complex bridge solution will be output.

[0059] Preferably, the loss function of the generative adversarial network of this solution uses binary cross entropy to measure the distribution difference between the generated data and the real data.

[0060] Furthermore, the corresponding parametric component family is called according to the bridge type scheme, and the parameters of the components in the parametric component family are adjusted according to the spatial coordinate point cloud data, such as adjusting the pier height parameter according to the point cloud elevation change; according to the above-mentioned topological connection logic, the component connection points are automatically located, and the called components are assembled, and the material properties, load labels and specification versions are automatically filled in the metadata; the IFC properties are automatically bound when the initial BIM bridge model is generated. After completion, the initial BIM bridge model is obtained, and the compliance of the initial BIM bridge model is checked.

[0061] Preferably, for connection types that cannot be assembled automatically (such as hinges of spatial curved surface components), the conflict position should be recorded in the log and a manual editing interface should be provided.

[0062] Optimally, a mathematical relationship matrix is ​​constructed between independent parameters (such as the total length of the bridge) and dependent parameters (such as the spacing between piers). This is then integrated into the BIM platform to embed dynamic parametric component families, enabling "one-stop modification, global linkage" and reducing repetitive modeling work. For non-standard bridge types (such as special-shaped cable-stayed bridges), generative adversarial networks are used in conjunction with graph convolutional networks to learn from historical data features to generate innovative bridge design solutions that conform to mechanical properties and topological logic. This improves the ability to model complex bridge topological relationships, enhances the structural rationality of complex bridge types, and effectively shortens complex design cycles. Automatically matching component positions based on point cloud data and embedding a code checking module ensures that the initial model meets design standards, reducing the risk of rework later.

[0063] S300: Performing lightweight finite element calculation on the initial BIM bridge model to obtain an optimized BIM bridge model.

[0064] Furthermore, lightweight finite element calculation is performed on the initial BIM bridge model to obtain an optimized BIM bridge model, including: geometric simplification and feature extraction of the initial BIM bridge model; adaptive mesh optimization of the simplified initial BIM bridge model to obtain a lightweight BIM bridge model; solver acceleration of the lightweight BIM bridge model and extraction of mechanical indicators; multi-objective optimization of the lightweight BIM bridge model after solver acceleration to obtain an optimized BIM bridge model.

[0065] Specifically, the geometric parameters of key components such as the main beam, pylons, and piers in the initial BIM bridge model were obtained, ignoring minor details such as bolt holes and decorative moldings. Using adaptive meshing technology, the mesh density was dynamically adjusted according to the stress gradient, and the initial BIM bridge model was adaptively meshed. High-stress areas were denser to 5mm, and low-stress areas were sparser to 50mm. When the stress gradient was greater than 50MPa / m, it was considered a high-stress area, and when the stress gradient was less than 10MPa / m, it was considered a low-stress area. When the stress gradient was 10MPa / m ≤ ≤ 50MPa / m, the mesh density was not adjusted. Unstructured mesh generation was also supported, and complex geometries (such as curved pylons and cable anchorage areas) were compatible, resulting in a lightweight BIM bridge model.

[0066] Furthermore, CUDA is used to accelerate the assembly and solution of the stiffness matrix of the lightweight BIM bridge model. A suitable sparse matrix compression storage format, such as the compressed sparse row (CSR) format, is adopted to store the non-zero elements in the stiffness matrix and their corresponding row and column indices in a one-dimensional array. Only the non-zero elements are stored to reduce memory usage. According to the compressed storage format, the matrix operation algorithm in the finite element analysis is adjusted. When performing operations such as matrix multiplication and solving linear equations, the compressed matrix is ​​directly operated to improve computational efficiency.

[0067] Furthermore, the various working conditions that need to be calculated (such as dead load + live load, extreme combinations of wind loads, etc.) are sorted and classified, and the load conditions and boundary conditions of each working condition are described and stored in detail. The multiple working conditions are combined into a batch task, the calculation order and related parameters of each working condition are defined, and a unique identifier is assigned to each working condition to facilitate the distinction and management of the calculation results. Multiple solver processes or threads are started at the same time to handle different working conditions respectively. Multi-threaded or distributed computing technology is used to solve different working conditions in parallel. During the calculation process, the calculation status and resource usage of each working condition are monitored and managed to ensure the smooth progress of the calculation. The results obtained from the calculation of each working condition are integrated and sorted, and the results are classified and stored according to the working condition identifier.

[0068] Furthermore, multi-objective optimization is performed on the lightweight BIM bridge model, including the definition of continuous variables and discrete variables. The continuous variables are continuously variable design parameters such as main beam height, pier diameter, and cable spacing, while the discrete variables are discrete design options such as section type and material type.

[0069] Furthermore, objective functions related to bridge modeling are constructed, including the objective function of minimizing the total mass of the structure and the objective function of maximizing stress uniformity, and the constraints related to bridge design are embedded in the lightweight BIM bridge model.

[0070] Preferably, lightweight helps to improve computing efficiency and shorten computing time.

[0071] Furthermore, the multi-objective optimization includes using the non-dominated sorting genetic algorithm II to randomly generate an initial population, where each individual represents a parameterized combination of bridge design schemes; calculating the multi-objective function value of each individual and performing non-dominated sorting; calculating the crowding distance of individuals in each non-dominated sorting level, and using the tournament selection method to select individuals from the current population as parent individuals; using a simulated binary crossover operator to perform a crossover operation on the selected parent individuals; using a polynomial mutation operator to perform a mutation operation on the offspring individuals, and checking whether the new individuals meet the relevant constraints of the bridge design.

[0072] Specifically, the NSGA-II (Non-dominated Sorting Genetic Algorithm II) algorithm was used to optimize the lightweight BIM bridge model. The population was initialized (randomly generating combinations of bridge design parameters). Each individual in the initial population represents a parameterized combination of bridge design options, including the values ​​of design variables such as girder height and pier diameter, as well as discrete parameters such as cross-section type and material type.

[0073] Furthermore, the individuals in the population are sorted by non-domination (dividing individuals into different levels). Non-dominated individuals (i.e., individuals that are not inferior to other individuals in all objective functions) are ranked first, and then non-dominated individuals are searched from the non-dominated individuals, and so on, until all individuals are ranked.

[0074] Furthermore, within each non-dominated level, the crowding distance of individuals is calculated (to maintain population diversity), representing the density of the objective function space surrounding the individual. Initially, border individuals (those at the extremes after objective function sorting) are assigned a larger crowding distance. Other individuals are then sorted by objective function value, and the crowding distance is calculated based on the difference in objective function values ​​between adjacent individuals. The crowding distance of an individual is the sum of its crowding distances across all objective function dimensions.

[0075] Furthermore, a tournament selection method (selecting parents) is used to select individuals from the current population as parents. During the selection process, individuals with higher non-dominated ranks are prioritized; if ranks are the same, individuals with higher crowding distances are selected to maintain population diversity.

[0076] Furthermore, a crossover operation is performed on the selected parent individuals (generating offspring individuals) to generate new individuals (offspring). A simulated binary crossover operator is used, with conventional crossover for continuous parameters such as main beam height and topologically constrained crossover for discrete parameters such as material combinations. Based on the crossover probability and distribution index, the corresponding genes of the parent individuals are recombined to generate new gene combinations, forming the genetic code of the offspring individuals. If the generated new individuals exceed the range of the design variables, boundary processing is performed and they are set to the boundary value.

[0077] Furthermore, mutation operations are performed on the offspring individuals (fine-tuning the offspring individuals). A polynomial mutation operator is used to make small random changes to the genes of the individuals based on the mutation probability and distribution index. The constraints are checked and the mutation rate of the current generation is calculated according to the adaptive mutation rate formula. The intensity of the mutation operation is adjusted based on this mutation rate to balance the global and local search capabilities. The adaptive mutation rate formula is as follows:

[0078] in, represents the mutation rate, represents the maximum mutation rate, t represents the current generation, and T represents the total generation.

[0079] Furthermore, for the newly generated individuals (offspring), check whether they meet the bridge design-related constraints. For individuals that violate the bridge design-related constraints, calculate their violation degree of the bridge design-related constraints, and penalize them according to the violation degree penalty function of the bridge design-related constraints to reduce their fitness.

[0080] Furthermore, the constraint breach penalty function formula is as follows:

[0081] in, represents the constraint penalty value, represents the Lagrange multiplier, P represents the parameter vector, i is the counting index, and n represents the total number of constraints. represents the i-th constraint function.

[0082] Furthermore, the parent population and the offspring population are merged to form a new population. The merged population is again subjected to non-dominated sorting and crowding distance calculation. Then, according to a certain selection strategy (such as retaining individuals of the previous level and retaining individuals with large crowding distances in the same level), a certain number of individuals are selected to form a new parent population (retaining individuals with high levels and high crowding distances).

[0083] Furthermore, the algorithm checks whether termination conditions are met, such as reaching the maximum number of iterations or the rate of change of the Pareto front being less than a given threshold. If the termination conditions are met, the algorithm stops and outputs the Pareto optimal solution set. Otherwise, the algorithm returns to the "crossover operation" and continues iterating until a Pareto optimal solution set that meets the termination conditions is obtained. One or more optimal solutions are selected from the Pareto optimal solution set, and the design variables of the selected optimal solutions are transferred back to the parametric component family. Since the components are dynamically linked through parameter relationship matrices and mathematical expressions (for example, the spacing between piers automatically adjusts with the total length of the bridge), the BIM model automatically reconstructs its geometry and properties after the parameters are updated.

[0084] Furthermore, a second finite element analysis (using adaptive meshing and CUDA acceleration) was performed on the multi-objective optimized BIM model to verify whether it met the structural performance requirements (such as stress, deformation, and stability). The optimized model was checked to see if it met the standards and design inputs in the specification documents. If any conflicts existed, the multi-objective optimization process was returned to adjust the relevant parameters of the optimized BIM model to obtain the optimized BIM bridge model.

[0085] Optimally, methods such as geometric simplification (ignoring minor details like bolt holes), adaptive meshing (stress gradient-driven meshing / sparseness), and CUDA acceleration can significantly reduce computational resource consumption and improve finite element analysis efficiency. The optimized NSGA-II algorithm balances objectives such as minimizing structural mass and maximizing stress uniformity. Incorporating a constraint violation penalty function ensures that the optimization solution achieves optimal safety, cost, and material quality. The optimization results are directly transferred back to the parametric component family, and the BIM model is automatically reconstructed through dynamic linkage, avoiding manual adjustment errors and accelerating the iterative closed loop.

[0086] S400: After performing code conflict detection and conflict parameter adjustment on the optimized BIM bridge model, a final bridge model is obtained.

[0087] Furthermore, after the optimized BIM bridge model is subjected to code conflict detection and conflict parameter adjustment, a final bridge model is obtained, including: the conflict detection includes geometric conflict detection and code conflict detection; the adjustment includes geometric adjustment, parameter correction and iterative verification.

[0088] Furthermore, the optimized BIM bridge model is checked for rules, including geometric collisions, load combinations and code requirements. A BIM platform plug-in is used to automatically scan conflict items in the model and generate a conflict report.

[0089] Furthermore, the conflict items in the conflict report are divided into geometric collision, load inconsistency, connection constraint violation, etc. according to the conflict type, and sorted according to safety, constructability and maintainability to determine the adjustment priority.

[0090] Furthermore, the parameter relationship matrix is ​​invoked to automatically locate conflicting parameters (such as beam height, pier span, and cable tension). Multiple candidate adjustment schemes are generated based on pre-set rules and optimization objectives. Lightweight finite element analysis and code verification are performed on each candidate adjustment scheme to evaluate the performance and code compliance of the adjusted model. The optimal adjustment scheme is selected using a decision matrix or weighted scoring method, integrating multiple objective indicators and conflict priorities. The parameters of the optimal adjustment scheme are written back into the parametric family, and the final bridge BIM model is regenerated and exported, along with a conflict resolution report and parameter modification log.

[0091] Preferably, use BIM plug-ins to scan geometric collisions and load combination conflicts, generate structured conflict reports, and reduce manual inspection costs; sort conflict items according to priorities such as safety and constructability, locate related parameters based on the parameter relationship matrix, generate multiple groups of adjustment plans and verify them through finite element analysis to ensure that the adjusted model takes into account both performance and specifications.

[0092] The present invention also provides a BIM-based bridge model intelligent construction system for implementing a BIM-based bridge model intelligent construction method. The system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. The processor executes the computer program to implement the BIM-based bridge model intelligent construction method.

[0093] An embodiment of the present invention provides a storage medium having a program stored thereon, which, when executed by a processor, implements the BIM-based bridge model intelligent construction method.

[0094] An embodiment of the present invention provides a processor, which is used to run a program, wherein the BIM-based bridge model intelligent construction method is executed when the program is run.

[0095] An embodiment of the present invention provides a device comprising a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, a method for intelligently constructing a bridge model based on BIM is implemented. The device herein may be a server, a PC, a PAD, a mobile phone, or the like.

[0096] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a BIM-based bridge model intelligent construction method.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may provide methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0102] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0103] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can be implemented using any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change RAM (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0105] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for intelligent construction of a bridge model based on BIM, characterized in that: include: collecting and preprocessing bridge model construction related data, and establishing a component library based on the preprocessed bridge model construction related data; defining geometric linkage rules and topological connection logic of components in the component library, and generating an initial BIM bridge model through a generative adversarial network; Performing lightweight finite element calculation on the initial BIM bridge model to obtain an optimized BIM bridge model; After code conflict detection and conflict parameter adjustment are performed on the optimized BIM bridge model, a final bridge model is obtained.

2. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: The bridge model construction related data includes: spatial coordinate data, design input data, historical bridge BIM model data and construction method case library data.

3. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: The definition of geometric linkage rules and topological connection logic of components in the component library includes: Extracting bridge parameters from historical bridge BIM model data and construction method case library data, and dividing the bridge parameters into independent parameters and dependent parameters; Establishing a parameter relationship matrix based on the independent parameters and the dependent parameters; Performing mathematical relationship conversion according to the parameter relationship matrix to obtain mathematical relationship expressions for various bridge components; The mathematical relationship formulas of the bridge components are embedded in the BIM platform to obtain a dynamically linked parametric component family.

4. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: The definition of geometric linkage rules and topological connection logic of components in the component library also includes: Classify the connection types of the components and prioritize the classified connection types. The priority ranking includes design priority, calculation priority, conflict resolution priority and maintenance priority; Modeling constraints according to the priority ranking to obtain component connection constraints; The component connection constraint conditions are embedded in the BIM platform, and dynamic connection verification is performed.

5. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: Generating an initial BIM bridge model by using a generative adversarial network includes: Constructing a training dataset for the generative adversarial network; Constructing a generator and a discriminator of the generative adversarial network; Designing a loss function and formulating a training strategy, and using the training dataset to train the generator and discriminator in stages according to the formulated training strategy; The bridge design parameters output by the generator are mapped and converted, and a bridge model is constructed using BIM to obtain an initial BIM bridge model.

6. The BIM-based bridge model intelligent construction method according to claim 5, characterized in that: The generator and discriminator of constructing the generative adversarial network include: A graph convolutional network is introduced into both the generator and the discriminator, and the graph convolutional network is used in the generator to encode the input graph structure data to obtain an embedded representation of the parameterized components and the connection types of the components; regenerate geometric parameters and topological structure of the bridge type according to the embedded representation; The discriminator processes the graph structure data of the bridge type through a graph convolutional network and learns the feature representation of parameterized components and the connection types of the components.

7. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: The performing lightweight finite element calculation on the initial BIM bridge model to obtain an optimized BIM bridge model includes: Perform geometric simplification and feature extraction on the initial BIM bridge model; Performing adaptive mesh optimization on the simplified initial BIM bridge model to obtain a lightweight BIM bridge model; Performing solver acceleration on the lightweight BIM bridge model and extracting mechanical indicators; Multi-objective optimization is performed on the lightweight BIM bridge model accelerated by the solver to obtain the optimized BIM bridge model.

8. The BIM-based bridge model intelligent construction method according to claim 7, characterized in that: The multi-objective optimization includes: The non-dominated sorting genetic algorithm II is used to randomly generate the initial population, and each individual represents a parameterized combination of bridge design options; Calculate the multi-objective function value of each individual and perform non-dominated sorting; Calculate the crowding distance of individuals in each non-dominated sorting level and use the tournament selection method to select individuals from the current population as parent individuals; A simulated binary crossover operator is used to perform a crossover operation on the selected parent individuals; The polynomial mutation operator is used to perform mutation operations on offspring individuals and check whether the new individuals meet the constraints.

9. The BIM-based bridge model intelligent construction method according to claim 1, characterized in that: After the optimized BIM bridge model is subjected to code conflict detection and conflict parameter adjustment, a final bridge model is obtained, including: the conflict detection includes geometric conflict detection and code conflict detection; the adjustment includes geometric adjustment, parameter correction and iterative verification.

10. A BIM-based bridge model intelligent construction system, characterized by: The system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the BIM-based bridge model intelligent construction method according to any one of claims 1 to 9.

Images