A metro assembled bridge model generation method based on multi-constraint topology assembly and parameter iteration

Abstract

This invention provides a method for generating metro prefabricated bridge models based on multi-constraint topology assembly and parameter iteration, including: line benchmark analysis and adaptive spatial skeleton generation; construction of a parametric family library of prefabricated components; automatic assembly of components based on the assembly topology matrix; assembly tolerance compensation and refined node modeling; multi-objective parameter iterative optimization; and hierarchical model output. This invention combines cubic NURBS spatial curve fitting, assembly topology matrix, tolerance compensation model, and least squares iterative algorithm to achieve automated and high-precision generation of metro prefabricated bridge models, while also supporting assembly simulation and accuracy detection.

Description

Technical Field

[0001] This invention belongs to the field of bridge engineering technology, and in particular relates to a method for generating a subway prefabricated bridge model based on multi-constraint topology assembly and parameter iteration. Background Technology

[0002] With the rapid development of urban rail transit, prefabricated elevated bridges for subways have become the mainstream bridge type in urban rail transit projects due to their significant advantages such as fast construction speed, less on-site wet work, high degree of component standardization, easy quality control, and green environmental protection. Digital twin and BIM technologies have deeply penetrated the entire life cycle of subway bridge design, construction, and operation and maintenance, and high-precision, simulable, and manufacturable three-dimensional digital models are the underlying foundation and core carrier for realizing the application of the above technologies.

[0003] The existing related technologies have the following technical shortcomings:

[0004] 1. Insufficient accuracy of line fitting: Existing technology 1 (CN110717213A) discloses a method and device for rapid generation of BIM construction model of railway bridge. Although it adopts horizontal and vertical curve fitting and assembly matrix technology, it is geared towards conventional railway bridges and does not adapt algorithms to complex spatial lines such as small radius curves, large gradient changes, and track superelevation gradients that are unique to subways. Traditional polyline skeleton or ordinary spline fitting will produce significant geometric distortion in subway curved bridges and variable gradient bridges, which cannot meet the stringent requirements of subway engineering for the accuracy of spatial lines down to the millimeter level.

[0005] 2. Lack of assembly logic: Existing technology 2 (CN121834994A) discloses a method for generating urban bridge models based on parametric rule-driven modeling. It adopts CGA rule-driven modeling, which achieves a certain degree of automation. However, CGA rules are essentially a one-way mapping of "rules → geometry". It lacks mathematical modeling of assembly sequence, component subordination relationship and splicing surface matching logic. It cannot realistically simulate the on-site prefabrication and assembly process of subway prefabricated bridges, and it cannot support the pre-assembly simulation and collision detection of prefabricated components.

[0006] 3. Lack of tolerance control mechanism: Although existing parametric modeling research on bridges based on CATIA has established a component family library, it has not solved the problem of tolerance accumulation after automatic assembly of components. It lacks a mathematical compensation model for splicing gaps and positioning deviations, resulting in a systematic deviation between the generated digital model and the actual state of the construction site, which cannot be directly used for construction guidance and accuracy acceptance.

[0007] 4. Lack of precision closed-loop optimization: Existing methods are mostly "one-time generation" modes, lacking a multi-objective iterative optimization mechanism for model coordinate deviation, linear residuals, and topology matching degree. This makes it impossible to eliminate the cumulative transmission of errors in each stage of the modeling process, and it is difficult to ensure that the overall accuracy of the model can stably reach the millimeter-level control standard of subway engineering.

[0008] 5. Low degree of specialization for subways: Existing general municipal bridge modeling tools have not been deeply customized for the special structures of subway prefabricated bridges, such as superelevated track structures, prestressed segmental beams, integrated prefabricated piers, and noise-reducing expansion joints. The component family library and engineering coding system lack adaptability to the subway industry.

[0009] To address the aforementioned issues, this invention specifically designs a method for generating prefabricated subway bridge models based on multi-constraint topology assembly and parameter iteration. Summary of the Invention

[0010] To address the shortcomings of existing technologies, this invention provides a method for generating metro prefabricated bridge models based on multi-constraint topology assembly and parameter iteration. By combining cubic NURBS curve fitting, assembly topology matrix, tolerance compensation model, and least squares iterative algorithm, it achieves automated and high-precision generation of metro prefabricated bridge models, while also supporting assembly simulation and accuracy detection.

[0011] The present invention achieves the above-mentioned technical objectives through the following technical means.

[0012] A method for generating a prefabricated subway bridge model based on multi-constraint topology assembly and parameter iteration includes the following process:

[0013] Step 1: Import subway line data and terrain elevation data, use cubic NURBS spatial curve fitting algorithm to fit discrete pile coordinates, generate a three-dimensional reference skeleton of the bridge center, and complete multi-level grid adaptive segmentation along the skeleton to divide the span control points, segment division points and pier positioning points.

[0014] Step 2: Build a dedicated parametric family library for the subway and bind geometric linkage calculation formulas to each component. At the same time, encode the components uniformly according to the four-level engineering coding rules of bridge type-span number-component type-segment number, and bind the coding to the number in the subsequent assembly topology matrix to achieve a strong association between the coding and the subsequent topology assembly.

[0015] Step 3: Construct the component assembly topology matrix, combine the skeleton curve coordinates, topology matrix parameters and track superelevation parameters to calculate the global positioning coordinates of each prefabricated component, and complete the batch orderly assembly of components according to the assembly priority;

[0016] Step 4: Construct an assembly tolerance compensation model, use tolerance compensation vectors to correct the component positioning coordinates and splice width, match the subway construction tolerance standards, and then perform local mesh subdivision modeling on key nodes.

[0017] Step 5: Construct the least squares multi-objective optimization objective function, iteratively correct the model parameters until the preset accuracy threshold or the maximum number of iterations is met, and complete the model accuracy optimization;

[0018] Step 6: Export the optimized bridge superstructure, substructure, node components, track auxiliary components, and terrain layer as multi-format scene assets, and simultaneously output the assembly topology table and tolerance inspection report.

[0019] Further, step 1 includes:

[0020] Step 1.1: Original data analysis; If formal subway design documents are available, extract the three-dimensional coordinates of each bridge pile, track curvature, track superelevation, longitudinal slope, and bridge clearance parameters; If no design documents are available, analyze satellite imagery and digital elevation model based on geographic information system, and obtain the initial pile coordinate sequence through coordinate inverse calculation algorithm;

[0021] Step 1.2: Cubic NURBS curve fitting:

[0022]

[0023] in, The fitted spatial curve of the bridge center. For each stake point weight factor, the straight line segment =1.0, curve segment =1.2, For cubic B-spline basis functions, For node parameters, Indicates the sequence number of each control point. This indicates the total number of all control points at each stake;

[0024] Step 1.3: Skeleton meshing and segmentation;

[0025] According to single span Calculate the segmentation interval , For the total length of the bridge, segments are completed along the fitted curve to generate a three-level skeleton grid reference: span control points: determine the start and end positions of each span of the bridge; segment division points: determine the division position of the precast segments within each span; pier positioning points: determine the center positioning coordinates of the lower piers.

[0026] Furthermore, in step 3, the topology matrix is ​​assembled. for:

[0027]

[0028] In the matrix, ( , , ) indicates the first Local coordinates of each component relative to the skeleton reference point For component subordinate type identifier, Assign assembly priority numbers, Number the spliced ​​surfaces. ;

[0029] The formula for calculating the global absolute coordinates of a component is:

[0030]

[0031] in, The azimuth angle of the tangent to the skeleton curve. The actual track superelevation value for the line, ( , , () represents the absolute coordinates of the component in global space. , , For NURBS skeleton curves at node parameters The X, Y, and Z coordinates of the location; numbered according to assembly priority during assembly. Assemble the components in ascending order, completing the assembly of the lower piers, upper beams, and node components sequentially, and number them according to the splice surfaces. Match the contact surfaces of adjacent components.

[0032] Furthermore, in step 4, the tolerance compensation vector for: The formula for compensating for the width of the splice seam is: ;in, The three-dimensional coordinate compensation components of the component are obtained by iterative optimization of multi-objective parameters in subsequent step 5. This is the allowable error during construction; To design the width of the splice seam;

[0033] The corrected coordinates of the component and the width of the splice joint are:

[0034]

[0035] in,( , , () represents the global absolute coordinates of the component after tolerance compensation correction; This indicates the corrected actual width of the seam.

[0036] Further, step 5 includes:

[0037] A multi-objective function incorporating coordinate deviation, linear residual, and topological matching degree is established. The least squares method is used for iterative optimization of the model parameters, with dual termination conditions to ensure the overall model accuracy reaches millimeter-level engineering requirements. The multi-objective optimization objective function is as follows:

[0038]

[0039] In the formula, , For the first The deviation between the model and the measured coordinates of each point is ( , , ) is the coordinate of the point calculated for the model. , , () represents the measured coordinates of the line; The residuals for the skeleton line fitting are obtained by comparing the interpolated coordinates of the NURBS curve with the theoretical coordinates of the original stake-by-stake control points, and then calculating using the root mean square error. , For the first Three-dimensional coordinates of the original line control points; For NURBS curves at nodal parameters Interpolated coordinates at the location, control point number range =0,1,2....,n, the total number of control points is n+1; For component topology matching degree; , These are the weighting coefficients;

[0040] The iteration termination condition is: the objective function ≤10 −6 Or the number of iterations reaches the preset maximum number of iterations.

[0041] The present invention has the following beneficial effects:

[0042] 1. High accuracy of line fitting: The bridge skeleton is generated by a cubic NURBS spatial curve fitting algorithm with adaptive weighted factor adjustment. By setting the weight factor of curve segment / straight line segment differentially (1.2 vs 1.0), the geometric distortion problem of traditional polyline skeleton in subway small-radius curved bridges and large-gradient variable slope bridges is solved. It perfectly adapts to the complex horizontal and vertical spatial alignment of subway and track superelevation constraints. The accuracy of line fitting is improved by an order of magnitude compared with traditional methods.

[0043] 2. Possesses complete assembly logic: The assembly topology matrix is ​​introduced to quantify the assembly sequence, subordinate relationships, and splicing rules of prefabricated components, which can realistically simulate the on-site assembly process of subway prefabricated bridges, support pre-assembly simulation of prefabricated components, and expand the application scenarios of the model.

[0044] 3. Adapt to subway assembly tolerance requirements: An assembly tolerance compensation model is added, which can automatically correct splicing gaps and component positioning deviations according to construction standards. The model is highly consistent with the construction site conditions, meeting the subway's millimeter-level tolerance control requirements.

[0045] 4. Controllable accuracy closed loop: Through the least squares multi-objective parameter iterative algorithm, the model coordinates, linearity, and topological relationships are iteratively optimized to eliminate the accumulation of modeling errors, and the overall accuracy of the model is stable and meets the standards.

[0046] 5. Balancing specialization and versatility: Constructing a dedicated prefabricated component family library for subways, adapting to unique structures such as segmental beams, prestressed components, and track bases; simultaneously outputting multi-format files, natively interfaced with digital twin platforms, GIS systems, and simulation software, with low data delivery difficulty.

[0047] 6. High degree of automation: The entire process is driven by algorithms and parameters, from skeleton generation, component assembly, tolerance correction to accuracy optimization, achieving automated operation, greatly reducing manual operation and improving modeling efficiency. Attached Figure Description

[0048] Figure 1 A flowchart of a method for generating prefabricated subway bridge models based on multi-constraint topology assembly and parameter iteration;

[0049] Figure 2 A logic diagram for multi-objective parameter iterative optimization. Detailed Implementation

[0050] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the scope of protection of the present invention is not limited thereto.

[0051] Example 1:

[0052] Reference Figure 1 As shown, the method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration according to the present invention includes the following process:

[0053] Step 1: Line Baseline Analysis and Adaptive Spatial Skeleton Generation: Import subway line data and terrain elevation data, and use a cubic NURBS spatial curve fitting algorithm to fit the discrete pile-by-pillar coordinates to generate a three-dimensional baseline skeleton for the bridge center. Then, perform multi-level mesh adaptive segmentation along the skeleton to divide the span control points, segmentation points, and pier positioning points. The specific process is as follows:

[0054] Step 1.1: Parsing the raw data;

[0055] If formal subway design documents are available, extract the three-dimensional coordinates of each bridge pile, the curvature of the line, the superelevation of the track, the longitudinal slope, and the bridge clearance parameters; if no design documents are available, analyze satellite imagery and digital elevation model (DEM) based on geographic information system (GIS) and obtain the initial pile-by-pile coordinate sequence through coordinate inverse calculation algorithm.

[0056] Step 1.2: Cubic NURBS space curve fitting, the fitting formula is:

[0057]

[0058] in, The fitted spatial curve of the bridge center. For each stake point weight factor, the straight line segment =1.0, curve segment =1.2, For cubic B-spline basis functions, For node parameters, Indicates the sequence number of each control point. This indicates the total number of all control points at each stake;

[0059] The recursive expression for the cubic B-spline basis function is:

[0060]

[0061] in, =3 is the order of the curve. It is a node vector.

[0062] Step 1.3: Skeleton meshing and segmentation;

[0063] According to single span Calculate the segmentation interval , For the total length of the bridge, segments are completed along the fitted curve to generate a three-level skeleton grid reference: span control points: determine the start and end positions of each span of the bridge; segment division points: determine the division position of the precast segments within each span; pier positioning points: determine the center positioning coordinates of the lower piers.

[0064] This invention analyzes parameters such as coordinates, curvature, superelevation, and elevation for both scenarios with and without design documents. It uses cubic NURBS curve fitting to discretize each pile point, eliminating errors in the polyline skeleton. Then, it uniformly divides the fitted curve according to the single-span length to form a multi-level skeleton mesh, which serves as the assembly benchmark for all components. Compared with traditional polyline skeletons or ordinary spline fitting, cubic NURBS curves have three major mathematical advantages: local support, higher-order continuity, and adjustable weight factors. They can accurately fit the complex horizontal and vertical alignment of subway lines, and the fitting accuracy of curve / straight line segments can be adaptively enhanced by adjusting the weight factors.

[0065] Step 2: Construction of the parametric family library for prefabricated components: Build a dedicated parametric family library for the metro, including upper segment components, lower pier components, connecting node components, and track auxiliary components. Bind geometric linkage calculation formulas to each component. At the same time, encode the components uniformly according to the four-level engineering coding rules of bridge type-span number-component type-segment number. The coding is also bound to the assembly topology number in Step 3 to achieve a strong association between the coding and subsequent topology assembly.

[0066] Preferably, a customized component family library for subway prefabricated bridges is added, distinct from general municipal bridge components, including specialized components such as track bases, prestressed components, noise-reducing expansion joints, and integrated prefabricated piers. Each component is configured with a cross-sectional linkage formula, enabling automatic changes in geometric dimensions according to track parameters. Simultaneously, the component code is associated with subsequent topology assembly, improving data linkage. For example, taking the common prefabricated segmental box girder in subways, its cross-sectional area linkage formula is:

[0067]

[0068] In the formula, The width of the top plate of the box girder. The height of the box girder The thickness of the top plate of the box girder. The angle of inclination of the web. Automatically adjusts and corrects according to the curvature parameters of the line; This refers to the cross-sectional area of ​​the precast segmental box girder.

[0069] Step 3: Automatic assembly of components based on the assembly topology matrix: Construct the component assembly topology matrix, combine the skeleton curve coordinates, topology matrix parameters, and track superelevation parameters to calculate the global positioning coordinates of each prefabricated component, and complete the batch orderly assembly of components according to assembly priority; details are as follows:

[0070] A two-dimensional assembly topology matrix is ​​constructed to quantify four major constraints: component location, dependency type, assembly sequence, and splicing surface. The global absolute coordinate calculation formula for components is derived, and precise positioning is achieved by integrating track superelevation parameters. The assembly sequence of piers, beams, and nodes is strictly followed according to assembly priority, recreating the on-site prefabrication and assembly process. Among these, the two-dimensional assembly topology matrix... The expression is:

[0071]

[0072] In the matrix, ( , , ) indicates the first Local coordinates of each component relative to the skeleton reference point For component subordinate type identifier, Assign assembly priority numbers, Number the spliced ​​surfaces. ;

[0073] The formula for calculating the global absolute coordinates of a component is:

[0074]

[0075] in, The azimuth angle of the tangent to the skeleton curve. The actual track superelevation value for the line, ( , , () represents the absolute coordinates of the component in global space. , , For NURBS skeleton curves at node parameters The X, Y, and Z coordinates of the location; numbered according to assembly priority during assembly. Assemble the components in ascending order, completing the assembly of the lower piers, upper beams, and node components sequentially, and number them according to the splice surfaces. Match the contact surfaces of adjacent components.

[0076] Step 4: Assembly tolerance compensation and node fine modeling: Construct an assembly tolerance compensation model, use tolerance compensation vectors to correct component positioning coordinates and splice width, match subway construction tolerance standards, and then perform local mesh subdivision modeling for key nodes such as wet joints, prestressed ducts, and beam-pier connection blocks.

[0077] Among them, the tolerance compensation vector for: ;

[0078] The formula for compensating for the width of the splice seam is: ;

[0079] In the formula, The three-dimensional coordinate compensation components of the component are obtained by iterative optimization of multi-objective parameters in step 5. To design the width of the splice seam, The allowable error during construction is 0~2mm.

[0080] The corrected coordinates and joint width expressions for the component are as follows:

[0081]

[0082] in,( , , () represents the global absolute coordinates of the component after tolerance compensation correction; This indicates the corrected actual width of the seam.

[0083] Step 5: Multi-objective parameter iterative optimization: such as Figure 2 As shown, a least-squares multi-objective optimization objective function is constructed, and the model parameters are iteratively corrected until a preset accuracy threshold or the maximum number of iterations is met, thus completing the model accuracy optimization; specifically as follows:

[0084] A multi-objective function incorporating coordinate deviation, linear residual, and topological matching degree is established. The least squares method is used for iterative optimization of the model parameters, with dual termination conditions to ensure the overall model accuracy reaches millimeter-level engineering requirements. The expression for the multi-objective optimization objective function is as follows:

[0085]

[0086] In the formula, , For the first The deviation between the model and the measured coordinates of each point is ( , , ) is the coordinate of the point calculated for the model. , , ( ) represents the measured coordinates of the line.

[0087] The residuals for the skeleton line fitting are obtained by comparing the interpolated coordinates of the NURBS curve with the theoretical coordinates of the original stake-by-stake control points, and then calculating using the root mean square error. , For the first Three-dimensional coordinates of each original control point; For NURBS curves at nodal parameters Interpolated coordinates at the location, control point number range =0,1,2....,n, the total number of control points is n+1; The component topology matching degree takes values ​​in the range [0,1]. , These are weighting coefficients, with values ​​of 0.6, 0.3, and 0.1 respectively.

[0088] The iteration termination condition is: the objective function ≤10 −6 The iteration count may reach the preset maximum iteration count Tmax, which is preferably set to Tmax=50 in this embodiment.

[0089] Step 6: Layered Model Output: The optimized bridge superstructure, substructure, node components, track auxiliary components, and terrain layer are exported as multi-format scene assets, and assembly topology tables and tolerance inspection reports are output simultaneously. This invention adopts a layered output mode, distinguishing between structural entities, terrain layers, and auxiliary components, while being compatible with mainstream 3D formats and GIS vector formats. It is suitable for various application scenarios such as digital twins, simulation platforms, and geographic information systems, and is accompanied by engineering reports to meet the data needs of construction, acceptance, and operation and maintenance.

[0090] Example 2:

[0091] This embodiment uses a six-span elevated prefabricated box girder bridge of a subway in a certain city as the modeling object. The bridge has a spatial curved shape (minimum curve radius R=350m), with three prefabricated segmental beams per span. The actual track superelevation value is [value missing]. =120mm, standard assembly tolerance 2mm, the model is generated using the method of this invention, and the specific implementation steps are as follows:

[0092] Step 1: Line reference analysis and adaptive generation of spatial skeleton

[0093] Step 1.1: Raw Data Analysis: Import the official design documents (CAD drawings + line parameter table) of the subway bridge and the regional digital elevation model (DEM, resolution 5m), extract the 3D coordinate sequence of each of the 6 spans of the bridge (a total of 186 point-by-points), the line curvature (minimum R=350m), and the actual track superelevation value. =120mm, longitudinal slope (maximum longitudinal slope 3.5%), bridge clearance and other parameters.

[0094] Step 1.2: Cubic NURBS spatial curve fitting: Set the weight factor for each stake point according to the route shape, and the curve segment =1.2, straight line segment =1.0, substitute into the NURBS fitting formula and the cubic B-spline basis function formula, and fit 186 discrete pile-by-pile points to obtain a continuous and smooth bridge center space curve. After fitting, the curve curvature is continuous and the tangent azimuth angle transitions smoothly.

[0095] Step 1.3: Skeleton meshing and segmentation: Calculate the total length of the bridge. Combined with single span Calculate the segmentation interval The model is segmented along the fitted curve, resulting in 6 span control points, 18 segment division points, and 7 pier positioning points, forming a three-level skeleton grid model.

[0096] Step 2: Construction of the parametric family library for prefabricated components

[0097] Based on the prefabricated bridge structure system for subways, a parameterized family library was constructed, including precast segmental box girders (18 units), precast integrated pier-cap beams (7 sets), wet joint formwork (12 units), prestressed anchor plates (36 units), noise reduction expansion joints (6 units), seismic blocks (14 units), and track bases (18 units). The width of the box girder top plate... =12.6m, box girder height =2.8m, box girder top plate thickness Web inclination angle Substitute parameters into the cross-sectional area linkage formula to realize automatic adjustment of cross-sectional dimensions according to track curvature; calculate the cantilever length of the cap beam based on the track width; encode all components according to the rule of bridge type-span number-component type-segment number (e.g., "box girder-03-segment-02"), and synchronously enter the coding information into the subsequent assembly topology matrix.

[0098] Step 3: Automatic assembly of components based on the assembly topology matrix;

[0099] An assembly topology matrix is ​​constructed for all prefabricated components of this bridge. (43×6 matrix) Define the pier assembly priority. =1, Segment beam assembly priority =2, Assembly priority of nodes and auxiliary components =3; Enter the local coordinates, subordinate type, and splice surface number of each component. Combine the skeleton curve coordinates, curve tangent azimuth angle, and track superelevation parameters, substitute them into the global coordinate calculation formula, and solve for the absolute spatial coordinates of all components. Assemble 7 sets of precast piers, 18 precast segment beams, and splice nodes in order of assembly priority from low to high, and align adjacent segments according to the splice surface number.

[0100] Step 4: Assembly tolerance compensation and detailed node modeling

[0101] Set the design seam width =20mm, construction tolerance (i.e., allowable construction error) =2mm, substituting into the splice seam width compensation formula, we obtain the final splice seam width Δδ=22mm; construct the tolerance compensation vector. The coordinates of all components and the width of the splice joints are corrected.

[0102] For the wet joints of the beam, prestressed ducts, and seismic blocks of the beam piers, a local mesh subdivision algorithm is used to refine the model mesh, accurately reconstruct the structure of the cast-in-place layer, tubular ducts, and irregular blocks, and complete the node modeling.

[0103] Step 5: Multi-objective parameter iterative optimization

[0104] Set weighting coefficients =0.6、 =0.3、 =0.1 Maximum number of iterations Tmax = 50, precision threshold 10 −6 The model coordinate deviation, linear fitting residual, and topological matching degree are substituted into the multi-objective optimization function, and iterative calculation is initiated. This embodiment involves a total of 12 iterations, and the final objective function is... =4.2×10 −7 <10 −6 Once the accuracy threshold requirement is met, the iteration stops, and the overall accuracy optimization of the model is completed.

[0105] Step 6: Model layered output

[0106] The optimized model was divided into five categories: superstructure beams, substructure piers, connecting node components, track auxiliary components, and terrain layers. These were exported as FBX, OBJ, and GIS vector files, respectively. Component parameter tables, assembly topology tables, and tolerance inspection reports were also exported. The entire dataset was integrated into the subway digital twin platform for design visualization, construction simulation, and daily operation and maintenance.

[0107] The embodiments described above are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration, characterized in that, The process includes the following: Step 1: Import subway line data and terrain elevation data, use cubic NURBS spatial curve fitting algorithm to fit discrete pile coordinates, generate a three-dimensional reference skeleton of the bridge center, and complete multi-level grid adaptive segmentation along the skeleton to divide the span control points, segment division points and pier positioning points. Step 2: Build a dedicated parametric family library for the subway and bind geometric linkage calculation formulas to each component. At the same time, encode the components uniformly according to the four-level engineering coding rules of bridge type-span number-component type-segment number, and bind the coding to the number in the subsequent assembly topology matrix to achieve a strong association between the coding and the subsequent topology assembly. Step 3: Construct the component assembly topology matrix, and calculate the global positioning coordinates of each prefabricated component by combining the skeleton curve coordinates, topology matrix parameters and track superelevation parameters. In subsequent actual assembly, the components are assembled in batches according to the assembly priority. Step 4: Construct an assembly tolerance compensation model, use tolerance compensation vectors to correct the component positioning coordinates and splice width, match the subway construction tolerance standards, and then perform local mesh subdivision modeling on key nodes. Step 5: Construct the least squares multi-objective optimization objective function, iteratively correct the model parameters until the preset accuracy threshold or the maximum number of iterations is met, and complete the model accuracy optimization; Step 6: Divide the optimized model into five categories: upper assembly beams, lower pier structures, connecting node components, track auxiliary components, and terrain layers. Export each category as a multi-format scene asset and simultaneously output component parameter tables, assembly topology tables, and tolerance inspection reports. The entire set of data is connected to the subway digital twin platform for design display, construction simulation, and daily operation and maintenance.

2. The method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration according to claim 1, characterized in that, Step 1 includes: Step 1.1: Original data analysis; If formal subway design documents are available, extract the three-dimensional coordinates of each bridge pile, track curvature, track superelevation, longitudinal slope, and bridge clearance parameters; If no design documents are available, analyze satellite imagery and digital elevation model based on geographic information system, and obtain the initial pile coordinate sequence through coordinate inverse calculation algorithm; Step 1.2: Cubic NURBS curve fitting: in, The fitted spatial curve of the bridge center. For each stake point weight factor, the straight line segment =1.0, curve segment =1.2, For cubic B-spline basis functions, For node parameters, Indicates the sequence number of each control point. This indicates the total number of all control points at each stake; Step 1.3: Skeleton meshing and segmentation; According to single span Calculate the segmentation interval , For the total length of the bridge, segments are completed along the fitted curve to generate a three-level skeleton grid reference: span control points: determine the start and end positions of each span of the bridge; segment division points: determine the division position of the precast segments within each span; pier positioning points: determine the center positioning coordinates of the lower piers.

3. The method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration according to claim 2, characterized in that, In step 3, the topology matrix is ​​assembled. for: In the matrix, ( , , ) indicates the first Local coordinates of each component relative to the skeleton reference point For component subordinate type identifier, Assign assembly priority numbers, Number the spliced ​​surfaces. ; The formula for calculating the global absolute coordinates of a component is: in, The azimuth angle of the tangent to the skeleton curve. The actual track superelevation value for the line, ( , , () represents the absolute coordinates of the component in global space. , , For NURBS skeleton curves at node parameters The X, Y, and Z coordinates of the location; numbered according to assembly priority during assembly. Assemble the components in ascending order, completing the assembly of the lower piers, upper beams, and node components sequentially, and number them according to the splice surfaces. Match the contact surfaces of adjacent components.

4. The method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration according to claim 3, characterized in that, In step 4, the tolerance compensation vector for: The formula for compensating for the width of the splice seam is: ;in, The three-dimensional coordinate compensation components of the component are obtained by iterative optimization of multi-objective parameters in subsequent step 5. This is the allowable error during construction; To design the width of the splice seam; The corrected coordinates of the component and the width of the splice joint are: in,( , , () represents the global absolute coordinates of the component after tolerance compensation correction; This indicates the corrected actual width of the seam.

5. The method for generating a metro prefabricated bridge model based on multi-constraint topology assembly and parameter iteration according to claim 4, characterized in that, Step 5 includes: A multi-objective function incorporating coordinate deviation, linear residual, and topological matching degree is established. The least squares method is used for iterative optimization of the model parameters, with dual termination conditions to ensure the overall model accuracy reaches millimeter-level engineering requirements. The multi-objective optimization objective function is as follows: In the formula, , For the first The deviation between the model and the measured coordinates of each point is ( , , ) is the coordinate of the point calculated for the model. , , () represents the measured coordinates of the line; The residuals for the skeleton line fitting are obtained by comparing the interpolated coordinates of the NURBS curve with the theoretical coordinates of the original stake-by-stake control points, and then calculating using the root mean square error. , For the first Three-dimensional coordinates of the original line control points; For NURBS curves at nodal parameters Interpolated coordinates at the location, control point number range =0,1,2....,n, the total number of control points is n+1; For component topology matching degree; , These are the weighting coefficients; The iteration termination condition is: the objective function ≤10 −6 Or the number of iterations reaches the preset maximum number of iterations.

Images