Bridge substructure modeling scheme generation method, device, equipment and medium

Abstract

The application discloses a bridge substructure modeling scheme generation method and device, equipment and medium, relates to the field of bridge engineering technology, and comprises the following steps: obtaining bridge superstructure design input parameters, constructing substructure design task information, generating design parameters by using a large language model reasoning, and then automatically generating a modeling script and performing structure response calculation. According to the calculation result, different modeling strategies are judged and executed, and finally a target modeling scheme meeting the requirements is intelligently generated, the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scene are realized, and the design efficiency and accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of bridge engineering, and in particular to a bridge substructure modeling scheme generation method, device, equipment and medium. BACKGROUND

[0002] With the continuous development of bridge construction, non-standard bridge types (such as interchange ramp bridges, split-flow end widening bridges, etc.) are increasingly widely used, and the design requirements of these bridge types are increasingly complex and personalized. The traditional design method mainly relies on the understanding and judgment of engineers on design specifications and experience rules, but when facing non-standard bridge types, the standard template is no longer applicable, and engineers need to reconfigure the component layout and adjust the reinforcement size according to the bridge width, structure type and other conditions, and complete the structure modeling and checking through modeling tools (such as OpenSees, Midas, etc.), which are mostly based on a large number of built-in design templates for condition matching, lack of understanding and reasoning ability for input structure parameters, and the structure design and modeling process lacks an effective feedback mechanism, which cannot automatically adjust the design parameters according to the structure response results, resulting in low iteration efficiency of the design process, which is difficult to meet the rapid design and optimization requirements of complex bridge types in engineering practice. This process not only consumes time and effort, but also requires a high level of experience for engineers. While large language models have made significant progress in natural language processing, their strong context learning ability and few-shot prompting mechanism enable them to understand and generate complex textual information. These features provide a new approach to solving the above problems. Large language models can generate structure design parameters that meet design specifications by learning from a large number of historical design samples and specification provisions, and automatically generate finite element modeling scripts. Therefore, there is an urgent need for a modeling scheme generation method based on large language models to generate design parameters and automatically model bridge substructures in non-standard bridge type scenarios, improving design efficiency and accuracy. SUMMARY

[0003] The main purpose of the present application is to provide a bridge substructure modeling scheme generation method, device, equipment and medium, aiming to solve the technical problem of how to generate design parameters and automatically model bridge substructures in non-standard bridge type scenarios, and improve design efficiency and accuracy.

[0004] To achieve the above purpose, the present application provides a bridge substructure modeling scheme generation method, comprising:

[0005] obtaining design input parameters of a bridge superstructure;

[0006] constructing design task information of a bridge substructure according to the design input parameters of the bridge superstructure;

[0007] generating design parameters of the bridge substructure by reasoning the design task information through a large language model.

[0008] inputting the design parameters into the large language model for analysis inference to generate a modeling script;

[0009] performing structural response calculation through the modeling script to obtain structural performance results;

[0010] judging according to the structural performance results to obtain a judgment result;

[0011] executing different modeling strategies according to the judgment result to generate a target bridge substructure modeling scheme;

[0012] The judgment result includes that the structural performance result does not meet the preset performance threshold and the structural performance result meets the preset performance threshold. The step of executing different modeling strategies according to the judgment result to generate a target bridge substructure modeling scheme includes:

[0013] If the judgment result is that the structural performance result does not meet the preset performance threshold, an optimization adjustment instruction is generated, and the optimization adjustment instruction is fed back to the large language model to adjust the design parameters to obtain target design parameters;

[0014] At the same time, the modeling script corresponding to the structural performance result is analyzed. If the analysis result of the modeling script does not converge, a script optimization instruction is generated and fed back to the large language model for constraint to obtain a target modeling script;

[0015] If the judgment result is that the structural performance result meets the preset performance threshold, the design parameters corresponding to the structural performance result are taken as target design parameters, and the modeling script corresponding to the structural performance result is taken as a target modeling script;

[0016] A target bridge substructure modeling scheme is generated according to the target design parameters and the target modeling script.

[0017] In an embodiment, the step of constructing a bridge substructure design task prompt information according to the design input parameters of the bridge superstructure includes:

[0018] obtaining historical bridge superstructure input parameters;

[0019] analyzing the bridge superstructure input parameters to obtain a bridge deck width, a next pier bridge deck distribution span, and a structure type;

[0020] According to the bridge deck width, the next pier bridge deck distribution span, and the structure type, an example pair is obtained by screening and converting from a historical design sample library according to a preset requirement, wherein the historical design sample library is established based on the historical bridge superstructure input parameters, and the example pair is a corresponding relationship between the bridge superstructure input parameters and the bridge substructure design parameters.

[0021] extracting design specification requirements related to the substructure design from a specification database;

[0022] splicing the example pair, the design specification requirements and the superstructure input parameters to generate design task prompt information of the bridge substructure.

[0023] In an embodiment, the step of obtaining the example pair based on the bridge deck width, the next pier deck distribution span and the structure type from the historical design sample library according to a preset requirement, includes:

[0024] filtering a first sample set with the same structure type and a difference less than a preset width threshold from the historical design sample library;

[0025] filtering a second sample set with the same structure type and a next pier deck distribution span greater than a preset span threshold from the historical design sample library;

[0026] converting the samples in the first sample set and the second sample set into an example pair in text form.

[0027] In an embodiment, the step of generating the design parameters of the bridge substructure by reasoning the design task information through the large language model, includes:

[0028] obtaining historical superstructure input parameters;

[0029] inputting the design task information into the large language model to obtain an association result of the bridge substructure design task information, the design specification requirements and the historical superstructure input parameters by using the context learning ability of the large language model;

[0030] obtaining a structure design rule based on the historical superstructure input parameters through the few-sample prompt mechanism of the large language model;

[0031] reasoning based on the association result and the structure design rule to generate the design parameters of the bridge substructure.

[0032] In an embodiment, the step of inputting the design parameters into the large language model for analysis and reasoning to generate a modeling script, includes:

[0033] calculating the spatial coordinates of each pier node according to the pier quantity and size parameters in the design parameters;

[0034] defining the cross-section geometric properties of each section of the bent cap according to the bent cap size parameters in the design parameters;

[0035] According to the material parameter in the design parameter, a material parameter of a structure material is obtained;

[0036] Based on the spatial coordinates, the cross-section geometric properties and the material parameter, node definition instructions and element connection instructions are generated;

[0037] According to the load parameter in the design parameter, load application instructions are generated;

[0038] According to the boundary condition parameter in the design parameter, setting instructions of structure constraint conditions are generated;

[0039] The node definition instructions, the element connection instructions, the load application instructions and the setting instructions are integrated and input into the large language model for inference to generate a modeling script of a bridge substructure.

[0040] In an embodiment, the step of performing structure response calculation through the modeling script to obtain structure performance results comprises:

[0041] A structure analysis model is constructed, and model dimensions and degrees of freedom are initialized and set;

[0042] A solver, convergence criteria and loading step size of the structure analysis model are set;

[0043] Material parameters, geometric dimensions, nodes, components and boundary conditions are obtained according to the node definition instructions, the element connection instructions and the setting instructions of constraint conditions in the modeling script;

[0044] According to the load application instructions in the modeling script, self-weight load, T-beam constant load reaction force and braking force are calculated and applied;

[0045] According to the material parameters and geometric dimensions, the nodes, components and boundary conditions, and the self-weight load, T-beam constant load reaction force and braking force, the structure analysis model is input for static analysis to obtain core response values;

[0046] Based on the core response values, structure performance results are obtained, which include the maximum bending moment, shear force of the key section of the cap beam, and the horizontal displacement of the top of the pier and the vertical reaction force at the bottom of the pier.

[0047] In an embodiment, the step of generating a target bridge substructure modeling scheme according to the target design parameters and the target modeling script comprises:

[0048] According to the target design parameters and the target modeling script, an initial modeling scheme of a bridge substructure is generated, which includes node arrangement, component definition, load application and boundary condition setting;

[0049] Correlate the modeling scheme with an actual construction scene to generate a scene correlation report, wherein the actual construction scene includes geological conditions, climate conditions, and traffic conditions of a construction site;

[0050] Generate a cost estimation report in combination with material market prices and construction costs according to the target design parameters and the target modeling script;

[0051] Perform a safety evaluation on the modeling scheme to generate a safety evaluation report;

[0052] Evaluate the influence of the modeling scheme on the surrounding environment to generate an environmental impact evaluation report;

[0053] Integrate the scene correlation report, the cost estimation report, the safety evaluation report, and the environmental impact evaluation report to generate a comprehensive evaluation report;

[0054] Perform multi-dimensional score calculation on the comprehensive evaluation report to obtain a comprehensive score;

[0055] When the comprehensive score exceeds a preset score threshold, the initial modeling scheme is taken as a target bridge substructure modeling scheme.

[0056] In addition, to achieve the above-mentioned purpose, the present application also provides a bridge substructure modeling scheme generation device, which comprises:

[0057] An acquisition module is configured to acquire design input parameters of a bridge superstructure;

[0058] A processing module is configured to construct design task information of a bridge substructure according to the design input parameters of the bridge superstructure;

[0059] An inference module is configured to perform inference on the design task information through a large language model to generate design parameters of the bridge substructure;

[0060] A script generation module is configured to input the design parameters into the large language model for analysis and inference to generate a modeling script;

[0061] A calculation module is configured to perform structural response calculation through the modeling script to obtain structural performance results;

[0062] A judgment module is configured to make a judgment according to the structural performance results to obtain a judgment result;

[0063] A result module is configured to execute different modeling strategies according to the judgment result to generate a target bridge substructure modeling scheme, and is further configured to generate an optimization adjustment instruction if the judgment result is that the structural performance result does not satisfy the preset performance threshold, feed back the optimization adjustment instruction to the large language model to adjust the design parameter, and obtain a target design parameter; meanwhile, the modeling script corresponding to the structural performance result is analyzed, and if the analysis result of the modeling script does not converge, a script optimization instruction is generated and fed back to the large language model for constraint to obtain a target modeling script; if the judgment result is that the structural performance result satisfies the preset performance threshold, the design parameter corresponding to the structural performance result is taken as the target design parameter and the modeling script corresponding to the structural performance result is taken as the target modeling script; and the target bridge substructure modeling scheme is generated according to the target design parameter and the target modeling script.

[0064] In addition, to achieve the above-mentioned purpose, the present application also provides a storage medium, which is a computer readable storage medium, and a computer program is stored on the storage medium. The computer program is executed by a processor to implement the steps of the bridge substructure modeling scheme generation method described above.

[0065] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program. The computer program is executed by a processor to implement the steps of the bridge substructure modeling scheme generation method described above.

[0066] The present application obtains bridge superstructure design input parameters, constructs substructure design task information, generates design parameters by using a large language model, and then generates a modeling script and performs structural response calculation. Different modeling strategies are judged and executed according to the calculation result, and finally a target modeling scheme that meets the requirements is generated, so as to realize the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scene, and improve the design efficiency and accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor.

[0068] Figure 1 The flowchart of the first embodiment of the bridge substructure modeling scheme generation method of the present application;

[0069] Figure 2 The flowchart of the second embodiment of the bridge substructure modeling scheme generation method of the present application;

[0070] Figure 3 A flowchart of a bridge substructure modeling scheme generation method according to a third embodiment of the present application;

[0071] Figure 4 A schematic diagram of the module structure of a bridge substructure modeling scheme generation device according to the present application;

[0072] Figure 5 A schematic diagram of the device structure of the hardware operating environment involved in the bridge substructure modeling scheme generation method according to the embodiment of the present application.

[0073] The object implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0074] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and do not limit the present application.

[0075] In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below with reference to the drawings and the specific embodiments.

[0076] With the continuous development of bridge construction, non-standard bridge types (such as interchanges, ramp bridges, and variable-width bridges at split-flow ends) are increasingly widely used, and the design requirements of these bridge types are increasingly complex and personalized. The traditional design method mainly relies on the understanding and judgment of engineers on design specifications and experience rules, but when facing non-standard bridge types, the standard template is no longer applicable, and engineers need to reconfigure the component layout and adjust the reinforcement size according to the bridge width, structure type and other conditions, and complete the structure modeling and checking through modeling tools (such as OpenSees, Midas, etc.), mostly based on a large number of built-in design templates for condition matching, lacking understanding and reasoning ability for input structure parameters, and the structure design and modeling process lacks effective feedback mechanism, unable to automatically adjust design parameters according to structure response results, resulting in low iteration efficiency of design process, difficult to meet the rapid design and optimization requirements of complex bridge types in engineering practice. This process not only consumes time and effort, but also requires a high level of experience for engineers. Therefore, there is an urgent need for a modeling scheme generation method to realize the modeling of bridge substructure in non-standard bridge type scenarios, improve design efficiency and accuracy.

[0077] Therefore, the application provides a bridge substructure modeling scheme generation method to solve the above problems. The main solution of the embodiment of the application is: obtaining design input parameters of a bridge superstructure; constructing design task information of the bridge substructure according to the design input parameters of the bridge superstructure; reasoning the design task information through a large language model to generate design parameters of the bridge substructure; combining and analyzing the design parameters to generate a modeling script; performing structure response calculation through the modeling script to obtain structure performance results; judging the structure performance results to obtain a judgment result; executing different modeling strategies according to the judgment result to generate a target bridge substructure modeling scheme.

[0078] Based on the above, the embodiment of the application further provides a bridge substructure modeling scheme generation method, which refers to Figure 1 , Figure 1 The flowchart of the bridge substructure modeling scheme generation method according to the first embodiment of the application is shown in FIG. 1. In this embodiment, the bridge substructure modeling scheme generation method includes steps S10-S70.

[0079] Step S10, obtaining design input parameters of a bridge superstructure.

[0080] It should be noted that the design input parameters include bridge deck width, next pier bridge deck width, structure type (such as simply supported T-beam, small box girder, etc.), standard span, design load level, geological condition category of the region, seismic fortification intensity, climate environment action level, and bridge deck transverse and longitudinal slope information. Meanwhile, component-level parameters such as main girder form, T-beam or box girder piece number, beam height, flange width, diaphragm arrangement spacing, etc. are collected. For bridge sections with linear changes, geometric feature data such as bridge axis curvature radius, horizontal curve radius, variable width transition section length, and widening method, etc. are also obtained. If multiple bridges or split-flow areas are involved, the distance between double bridges, connection structure form, and traffic organization characteristics need to be identified. All parameters are obtained through analyzing design drawings, importing BIM model data, reading database records, or inputting through a human-computer interaction interface, etc. and are processed for data cleaning, unit unification, and format standardization to ensure the completeness, accuracy, and consistency of the input information.

[0081] Step S20, constructing design task information of the bridge substructure according to the design input parameters of the bridge superstructure.

[0082] It should be noted that this step converts complex engineering design problems into language tasks that can be understood and processed by a large language model. This process is not just a simple concatenation of input parameters, but also a “hint engineering” process.

[0083] Further, step S20 also includes: first, obtaining historical bridge superstructure input parameters, specifically, the input parameters of the superstructure need to be obtained from historical bridge design projects, these parameters include but are not limited to bridge deck width, structure type, span distribution and other key information. Through the analysis of these historical data, the bridge deck width, next pier bridge deck distribution span and structure type and other specific parameters required by the current design task can be extracted. These parameters are the basis for designing the bridge substructure, because they directly affect the design and layout of the bridge pier, bent cap and foundation and other substructures.

[0084] Next, the bridge superstructure input parameters are analyzed to obtain the bridge deck width, the next pier deck distribution span, and the structure type. Then, based on the bridge deck width, the next pier deck distribution span, and the structure type, the example pairs are selected and converted from the historical design sample library according to the preset requirements. The specific steps of the example pairs are as follows: selecting a first sample set from the historical design sample library that has the same structure type and a bridge deck width difference less than a preset width threshold; selecting a second sample set from the historical design sample library that has the same structure type and a next pier deck distribution span greater than a preset span threshold; and converting the samples in the first sample set and the second sample set into text form example pairs, which are the corresponding relationship between the bridge superstructure input parameters and the bridge substructure design parameters. Specifically, a first sample set is selected from the historical design sample library that has the same structure type and a bridge deck width difference less than a preset width threshold. This selection process ensures that the selected samples have high similarity to the current design task in terms of bridge deck width, thereby providing a more practical reference for subsequent design. For example, if the bridge deck width of the current design task is 20 meters and the preset width threshold is 2 meters, the bridge deck width of the selected first sample set should be between 18 meters and 22 meters. Then, a second sample set is selected from the historical design sample library that has the same structure type and a next pier deck distribution span greater than a preset span threshold. Finally, the samples in the first sample set and the second sample set are converted into text form example pairs. These example pairs not only contain key information such as sample bridge deck width, structure type, and span distribution, but also include related substructure design parameters such as the number of piers, cap beam size, and foundation type. By converting these samples into text form, they can be more easily incorporated into the design task prompt information, providing rich context information for large language models to more accurately infer the required substructure design parameters for the current design task. The historical design sample library mentioned above is established based on historical bridge superstructure input parameters and contains various bridge design cases under different conditions, which provide rich references for current design. By selecting similar historical samples from the current design task, previous successful experiences can be learned to improve design efficiency and quality. Then, the design specification requirements related to the substructure design are extracted from the specification database. Specifically, the design specification requirements related to the substructure design are extracted from the specification database. These specification requirements are the guidelines that must be followed during the design process, ensuring the safety, reliability, and durability of the bridge. Design specifications cover multiple aspects such as material selection, structure size, and load standards, and are an important basis for bridge design.

[0085] Finally, the example pairs, design specification requirements, and superstructure input parameters are spliced to generate the design task prompt information of the bridge substructure. Specifically, the selected example pairs, extracted design specification requirements, and current superstructure input parameters are spliced to generate the design task prompt information of the bridge substructure. This prompt information provides clear guidance for subsequent design work, as it not only includes historical design experience, but also combines specific requirements and specification standards of the current design. In this way, designers can quickly generate substructure design schemes that meet the requirements, while ensuring the rationality and standardization of the design scheme. This design method based on historical data and specification requirements not only improves design efficiency, but also reduces design risk, providing a solid guarantee for the successful implementation of bridge engineering.

[0086] Step S30, reasoning the design task information by the large language model to generate the design parameters of the bridge substructure.

[0087] It should be noted that the large language model, through its powerful context learning ability and few-shot prompting mechanism, can understand and process complex design task information, thereby generating design parameters that meet engineering requirements.

[0088] Specifically, the design task information is input into the large language model, which is an artificial intelligence technology based on deep learning, designed specifically to understand and generate natural language text. These models are trained on a large amount of text data to learn the patterns and structures of language, enabling them to perform various natural language processing tasks such as text generation, translation, question answering, etc. Instead of simply looking up or matching answers, the model uses its powerful In-Context Learning mechanism and Few-shot Learning ability to deeply understand and logically reason the information. Specifically, the model will first identify and analyze the key elements in the prompt information, such as the current bridge width, structure type, etc. Then, it will analyze the provided historical samples to induce potential relationships and variation rules between different design conditions and substructure parameters. For example, the model may learn that as the bridge width increases, the size of the cap beam and the number of piers also need to be adjusted accordingly. At the same time, the model will consider design specifications as hard constraints to ensure that the parameters it reasons out do not violate basic engineering guidelines. Finally, through this multi-dimensional, cross-domain comprehensive reasoning, the large language model can generate a complete set of structured design parameters, including pier number and spacing, cap beam size, and foundation type, etc. These parameters are not only based on direct derivation of input, but also incorporate the model's understanding of historical experience and specification requirements, thus achieving intelligent conversion from natural language tasks to professional engineering data.

[0089] Step S40, input the design parameters into the large language model for analysis and reasoning, and generate a modeling script.

[0090] It should be noted that according to the material properties and cross-sectional dimensions provided in the design parameters, the physical properties of each material and the geometric parameters of each cross-section are defined in detail, which will directly affect the mechanical behavior of the structure and the analysis results. After defining the material and cross-section properties, the coordinates of each node need to be calculated accurately according to the size and arrangement of the pier, bent cap and foundation in the design parameters, and numbered according to certain rules. Then, according to the connection relationship of the nodes, define the structural units such as column units of the pier and beam units of the bent cap, etc., to ensure that the geometric shape of the structure model is completely consistent with the design intention. In addition, according to the design parameters and the requirements of structural analysis, appropriate boundary conditions and loads need to be applied, including self-weight, vehicle load, wind load, earthquake load, etc., and the accurate application of these loads is the key to simulating the actual working condition. Finally, set the type and corresponding control parameters of the structural analysis, such as loading step, convergence criterion, etc., which control the process and precision of the structural analysis. All these definitions and instructions are integrated into a complete modeling script, output in a code format that can be recognized by the structural analysis software.

[0091] Specifically, step S40 includes: first, according to the number of piers and size parameters in the design parameters, the spatial coordinates of each pier node are calculated, which not only includes the planar position of the pier, but also includes its height in the vertical direction, ensuring the accurate positioning of the pier in three-dimensional space. Then, according to the cap beam size parameters in the design parameters, the cross-sectional geometric properties of each section of the cap beam are defined, such as width, height and cross-sectional shape, etc. These properties directly affect the mechanical properties and structural behavior of the cap beam. Next, according to the material parameters in the design parameters, the material parameters of the structural material are set, such as elastic modulus, Poisson's ratio, density, etc. These parameters are the basis for calculating internal forces and deformations in structural analysis. Subsequently, based on the spatial coordinates, cross-sectional geometric properties and material parameters, node definition instructions and element connection instructions are generated. Node definition instructions describe the location and number of each node in detail, while element connection instructions define the connection relationship between nodes, forming the overall framework of the structure. Then, according to the load parameters in the design parameters, load application instructions are generated, including dead load, vehicle load, wind load, seismic load, etc. Accurate application of loads is the key to simulating actual working conditions. According to the boundary condition parameters in the design parameters, setting instructions for structural constraint conditions are generated, which usually include fixed constraints, angular constraints and displacement constraints. These conditions ensure the stability and safety of the structure in actual use. Finally, the node definition instructions, element connection instructions, load application instructions and setting instructions are integrated and input into the large language model for reasoning, generating the modeling script of the bridge substructure. Specifically, after integrating the above instructions, they are input into the large language model, which can generate modeling scripts that conform to the syntax of specific structural analysis software (such as OpenSeesPy) using its powerful reasoning ability. Through context learning and few-shot prompting mechanism, the large language model can understand the meaning of these instructions and generate accurate modeling scripts. The generated script not only includes node definition, element connection, load application and boundary condition setting instructions, but also can automatically adjust these instructions according to the design parameters, ensuring the integrity and accuracy of the model. The final script contains all necessary information from material definition to load application and can be directly used for structural analysis. Through this process, design efficiency can be significantly improved, human errors can be reduced, and the accuracy and consistency of the model can be ensured. This method is not only suitable for standard bridge types, but also can effectively handle non-standard bridge types and bridge design under complex environments, providing strong support for the rapid construction and safe operation of bridges.

[0092] Step S50, structural response calculation is performed through the modeling script to obtain the structural performance results.

[0093] It is worth noting that the structural response calculation through the modeling script is a key step to verify whether the design scheme meets the engineering requirements. This process involves importing the modeling script into structural analysis software such as OpenSeesPy, Midas, etc. for detailed structural analysis. The modeling script contains all the necessary information of the bridge substructure, including node coordinates, element connections, material properties, boundary conditions, and load application, etc. After these information are read by the structural analysis software, the software will calculate according to the preset analysis type (such as static analysis, modal analysis, transient analysis, etc.).

[0094] During the calculation process, the software will gradually apply loads and calculate the response of the structure according to the material and geometric properties. These responses include node displacement, internal force distribution, stress and strain, etc. Key indicators. For example, in static analysis, the software will calculate the deformation and internal force distribution of the structure under the action of dead load and live load; in modal analysis, the software will determine the natural frequency and mode shape of the structure; in transient analysis, the software will simulate the response of the structure under dynamic load. After the calculation is completed, the structural analysis software will output detailed structural performance results. These results are usually presented in the form of charts, numerical tables or animations, which are convenient for engineers to evaluate. By analyzing these results, engineers can judge whether the structure meets the requirements of the design specification, such as whether the maximum displacement is within the allowable range, whether the stress of the key section is lower than the material strength, etc. If the structural performance does not meet the requirements, engineers can adjust the design according to these results, such as modifying the section size, adjusting the material properties or rearranging the components, etc.

[0095] Through the structural response calculation by the modeling script, not only the efficiency of design verification is improved, but also the accuracy and reliability of the design scheme are ensured. This process is an indispensable part of modern bridge design, which enables engineers to comprehensively evaluate and optimize the design scheme before construction, thereby improving the safety, durability and economy of the bridge.

[0096] Step S60, judging according to the structural performance results to obtain a judgment result.

[0097] It is worth noting that this process involves evaluating the structural performance results output by the structural analysis software, which usually includes node displacement, internal force distribution, stress and strain, natural frequency, mode shape, etc. Key indicators. These indicators reflect the response of the structure under different load conditions and are an important basis for evaluating the safety and reliability of the structure.

[0098] Specifically, engineers need to compare the structural performance results with the requirements of design specifications item by item. For example, for the joint displacement, it needs to be determined whether it is within the allowed range; for the stress of the key section, it needs to be confirmed whether it is lower than the strength limit of the material; for the natural frequency, it needs to be ensured that it is out of sync with the possible external excitation frequency to avoid resonance phenomenon. These criteria are usually clearly given by national or industry standard specifications, which are the guidelines that must be followed in the design process.

[0099] When making judgments, multiple factors need to be considered comprehensively. For example, for the joint displacement, not only the absolute value needs to be considered whether it is within the allowed range, but also the influence of the displacement on the overall stability of the structure. For the internal force distribution, it needs to be checked whether there is local stress concentration phenomenon, which may lead to local damage of the structure. For the stress and strain results, it needs to be evaluated in combination with the fatigue performance of the material, especially whether the material can withstand repeated load action without fatigue failure within the service life of the bridge.

[0100] Step S70, according to the judgment result, different modeling strategies are executed to generate the target bridge substructure modeling scheme.

[0101] It should be noted that this process not only needs to analyze the preliminary structural performance results in detail, but also needs to select appropriate modeling strategies according to these results to generate the target modeling scheme that meets the design requirements.

[0102] Specifically, if the judgment result is that the structural performance result does not meet the preset performance threshold, an optimization adjustment instruction is generated, indicating that there are some problems in the current design and optimization adjustment is needed. At this time, the optimization adjustment instruction is generated, which usually includes specific adjustment suggestions for design parameters, such as increasing the size of the component, changing the material properties, adjusting the node position, etc. The optimization adjustment instruction is fed back to the large language model to adjust the design parameters to obtain the target design parameters. These adjustment instructions are then fed back to the large language model for targeted adjustment of the design parameters. The large language model uses its powerful context learning ability and few-shot prompting mechanism to intelligently optimize the design parameters in combination with historical design samples and specification requirements. For example, if the analysis result shows that the bending moment of the key section of the cap beam exceeds the allowed range, the large language model may suggest increasing the section height of the cap beam or adjusting the material properties to improve its carrying capacity. At the same time, the modeling script corresponding to the structural performance result is analyzed, and if the analysis result of the modeling script does not converge, a script optimization instruction is generated and fed back to the large language model for constraint to obtain the target modeling script. Specifically, if the analysis result of the modeling script does not converge, i.e., there are syntax errors or logical problems, the system will generate a script optimization instruction. These instructions are also fed back to the large language model for constraint and optimization of the modeling script. The large language model will adjust the node definition, unit connection, load application, and boundary condition setting in the script according to these instructions to generate a new target modeling script. This process ensures the grammatical correctness and execution efficiency of the modeling script, avoiding structural analysis failure due to script errors. If the judgment result is that the structural performance result meets the preset performance threshold, the design parameters corresponding to the structural performance result are taken as the target design parameters and the modeling script corresponding to the structural performance result is taken as the target modeling script.

[0103] Finally, the target bridge substructure modeling scheme is generated according to the target design parameters and the target modeling script. However, various factors need to be considered when generating the target modeling scheme. Therefore, an initial modeling scheme of the bridge substructure is first generated according to the target design parameters and the target modeling script. The initial modeling scheme includes node arrangement, component definition, load application, and boundary condition setting. Then, the modeling scheme is associated with the actual construction scene to generate a scene association report. The actual construction scene includes the geological conditions, climate conditions, and traffic conditions of the construction site. Geological conditions directly affect the selection of foundation design and construction methods, such as special reinforcement measures may be needed for construction on soft soil foundation. Climate conditions affect construction progress and material selection, such as waterproof measures may need to be considered for construction in rainy areas. Traffic conditions affect traffic organization and construction safety during construction, such as phased construction strategies may need to be adopted for construction on busy roads. According to the target design parameters and the target modeling script, combined with material market prices and construction costs, a cost estimation report is generated. Cost estimation not only includes material costs, but also covers construction equipment, labor costs, transportation costs, etc. Accurate cost estimation helps to develop project budget and cost control, ensuring the economic feasibility of the project. A safety evaluation report is generated by evaluating the safety risks and operation difficulties during construction, including structural stability, safety of construction equipment, safety measures for construction personnel, etc. Through safety evaluation, potential safety hazards can be identified and precautions can be taken in advance to ensure the safety of the construction process. An environmental impact assessment report is generated by evaluating the impact of the modeling scheme on the surrounding environment, including noise, dust, water pollution, etc. Through environmental impact assessment, appropriate environmental protection measures can be taken to reduce the negative impact of construction on the environment and ensure the environmental friendliness of the project. Then, a comprehensive evaluation report is generated by integrating the scene association report, cost estimation report, safety evaluation report, and environmental impact assessment report. The comprehensive evaluation report provides a comprehensive perspective on the various aspects of the modeling scheme. A multi-dimensional score calculation is performed on the comprehensive evaluation report to obtain a comprehensive score. The scoring dimensions can include economy, safety, environmental friendliness, construction feasibility, etc. Finally, a multi-dimensional score calculation is performed on the comprehensive evaluation report to obtain a comprehensive score. When the comprehensive score exceeds the pre-set score threshold, it indicates that the modeling scheme has reached a high standard in various aspects. The initial modeling scheme is taken as the target bridge substructure modeling scheme.

[0104] Through this judgment result-based modeling strategy selection and iterative optimization process, a high-quality target bridge substructure modeling scheme can be generated. This method not only improves the accuracy and reliability of the design, but also ensures the feasibility and economy of the design scheme, providing strong support for the rapid construction and safe operation of the bridge.

[0105] The embodiment obtains bridge superstructure design input parameters, constructs substructure design task information, generates design parameters by using a large language model reasoning, and then generates modeling scripts and performs structure response calculation. According to the calculation result, different modeling strategies are judged and executed, and finally the target modeling scheme meeting the requirements is generated, realizing the design parameter generation and automatic modeling of the bridge substructure in the non-standard bridge type scene, and improving the design efficiency and accuracy.

[0106] Based on the first embodiment of the present application, the same or similar contents as the above embodiment one can be referred to the above introduction, and the following will not be repeated. On this basis, please refer to Figure 2 , the bridge substructure modeling scheme generation method step S30 further comprises steps S201-S204:

[0107] Step S201, obtaining historical superstructure input parameters.

[0108] It should be noted that the historical superstructure input parameters include bridge deck width, structure type, span distribution, load condition, environmental factors, and construction technology and economic cost. By collecting and sorting these historical data, rich reference information can be provided for the current design task. For example, the bridge deck width directly affects the size and arrangement of the pier and bent cap, and the structure type determines the mechanical behavior and stress characteristics of the substructure. Different types of bridges, such as simply supported beam bridges, continuous beam bridges, cable-stayed bridges, etc., have significant differences in the requirements for the substructure. The size of the structure span not only affects the height and spacing of the pier, but also determines the type and size of the foundation. Large-span bridges usually need to adopt deep foundation and high pier design to ensure the stability and carrying capacity of the structure. Accurate assessment of load conditions is crucial for the resistance design of the substructure, and the combined effect of various loads such as dead load, live load, wind load, and seismic load needs to be considered. Environmental factors such as geological conditions and hydrological conditions directly affect the selection of foundation type and the development of construction scheme. The consideration of construction technology and economic cost ensures the implementability and economic rationality of the design scheme.

[0109] Step S202, inputting the design task information to the large language model

[0110] It should be noted that the large language model can understand and process complex design task information through its powerful context learning ability, thereby generating results associated with bridge lower structure design task information, design specification requirements, and historical upper structure input parameters. Specifically, the design task information includes the bridge deck width, structure type, span distribution, and other key parameters of the current bridge upper structure, which are the basis for designing the bridge lower structure. At the same time, the design specification requirements related to the lower structure design are extracted from the specification database, which are the guidelines that must be followed in the design process to ensure the safety, reliability, and durability of the bridge. In addition, historical upper structure input parameters provide rich reference information, which can help the large language model better understand and generate design parameters related to the current design task.

[0111] The large language model can identify and apply design rules and patterns by analyzing these input information, generating lower structure design parameters that match the current design task. For example, the model can infer appropriate pier spacing and size based on the bridge deck width and structure type to ensure the stability and load-bearing capacity of the structure. At the same time, the model will also consider the requirements of the design specification to ensure that the generated design parameters meet industry standards and safety specifications. In this way, the large language model not only generates design parameters that meet the requirements, but also provides results associated with design task information, design specification requirements, and historical upper structure input parameters, providing comprehensive guidance for subsequent design work.

[0112] Step S203, based on the historical upper structure input parameters, the structure design law is obtained through the few-shot prompting mechanism of the large language model.

[0113] It should be noted that by providing a small number of representative samples, the large language model is guided to understand and generate design parameters and schemes related to the current design task. Specifically, the few-shot prompting mechanism uses key samples from historical data, which contain important parameters such as bridge deck width, structure type, span distribution, and corresponding lower structure design parameters such as pier size, cap beam size, and foundation type. By analyzing these samples, the large language model can learn the internal relationships between different parameters and design rules.

[0114] For example, the model can identify the reasonable spacing and size of the pier, as well as the appropriate cross-sectional size of the cap beam, under specific bridge deck width and structure type. This learning process not only relies on sample data, but also combines design specifications and engineering experience to ensure that the generated design parameters not only meet the specification requirements, but also have practical engineering feasibility. Through the few-shot prompting mechanism, the large language model can quickly generate high-quality design schemes based on limited samples, improving design efficiency and accuracy.

[0115] Step S204, reasoning based on the association results and structural design rules to generate the design parameters of the bridge substructure.

[0116] It should be noted that this process fully utilizes the context learning ability and few-shot prompting mechanism of the large language model. By analyzing historical data and current design task information, it extracts rules and patterns related to bridge substructure design. Based on the above association results and design rules, the large language model reasons and generates the design parameters of the bridge substructure. For example, the model can infer the appropriate pier spacing and size based on the current bridge deck width and structure type, combined with the design rules in historical data. At the same time, the model will also consider the requirements of design specifications to ensure that the generated design parameters meet industry standards and safety specifications. In this way, the large language model not only generates design parameters that meet the requirements, but also provides results associated with design task information, design specification requirements, and historical substructure input parameters, providing comprehensive guidance for subsequent design work.

[0117] This embodiment obtains historical substructure input parameters, then inputs design task information into the large language model, and uses its context learning ability to obtain the association results of design task information, design specification requirements, and historical parameters. Then, through the few-shot prompting mechanism, it extracts structural design rules from historical parameters. Finally, based on these association results and design rules, it reasons and generates the design parameters of the bridge substructure. Through the context learning and few-shot prompting mechanism of the large language model, it quickly generates substructure parameters that meet the design specifications, improving design efficiency and accuracy, especially suitable for complex bridge types and non-standard design scenarios, providing intelligent design support for bridge engineering.

[0118] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as the above embodiment one can refer to the above introduction, and the following will not be repeated. On this basis, please refer to Figure 3 , the bridge substructure modeling scheme generation step S50 further includes steps S301-S306:

[0119] Step S301, construct a structural analysis model, initialize the model dimension and degree of freedom.

[0120] It should be noted that the dimension of the model determines the complexity and accuracy of the analysis. For a simple pier model, a two-dimensional model is usually used, which can simplify the calculation and improve the efficiency. In this case, the dimension of the model is set to two-dimensional, and the degree of freedom usually includes three-direction translational freedom. However, for more complex bridge structures such as cable-stayed bridges or suspension bridges, a three-dimensional model may be necessary to accurately capture the behavior of the structure in space. In this case, the dimension of the model is set to three-dimensional, and the degree of freedom includes six-direction translational and rotational freedom.

[0121] Step S302, set the solver of the structure analysis model, the convergence criterion and the loading step.

[0122] It should be noted that the selection of the solver directly affects the calculation efficiency and accuracy. Common solvers include direct methods and iterative methods. Direct methods such as Gaussian elimination are suitable for small or medium-sized linear problems, and the calculation results are accurate, but the calculation cost increases significantly as the problem size increases. Iterative methods such as the conjugate gradient method are suitable for handling large-scale sparse matrix problems and have high computational efficiency, but may require more iterations to achieve the required accuracy. For nonlinear problems, such as bridge analysis considering material nonlinearity or geometric nonlinearity, iterative solvers such as the Newton-Raphson method are usually used to handle complex nonlinear relationships by gradually approaching the true solution.

[0123] The convergence criterion is a standard for determining whether the iterative solution process has reached a stable solution. Common convergence criteria include displacement convergence criteria and force convergence criteria. Displacement convergence criteria compare the node displacement changes between consecutive iterations to determine whether convergence has been achieved. When the displacement change is less than a predetermined threshold, the solution is considered to have converged. Force convergence criteria check whether the unbalanced force on the node is less than a predetermined threshold to determine convergence. In actual analysis, both criteria are usually used together to ensure the stability and accuracy of the solution.

[0124] The loading step determines the size of the load increment. For linear problems, a larger loading step can usually be used because the response of a linear problem is proportional to the load, and the calculation process is relatively simple. However, for nonlinear problems such as bridge analysis considering large deformation or material nonlinearity, a smaller loading step is required to ensure that the incremental response of each step can be accurately captured. Although a smaller loading step increases the amount of calculation, it effectively improves the accuracy and stability of the solution of nonlinear problems. In actual operation, engineers need to adjust the loading step flexibly according to the specific analysis requirements and problem characteristics to balance the calculation efficiency and result accuracy. By reasonably setting the solver, convergence criterion and loading step, the efficiency and reliability of the structure analysis can be effectively improved, providing accurate mechanical basis for bridge design.

[0125] Step S303, according to the node definition instructions, element connection instructions and constraint condition setting instructions in the modeling script, get material parameters, geometric dimensions, nodes, components and boundary conditions.

[0126] It is important to note that the node definition instructions provide detailed information about the spatial location and numbering of each node, which serves as the foundation for constructing the structural model. By parsing these instructions, the coordinates of each node can be determined, providing accurate reference points for subsequent member connections and load application. For example, in a bridge pier model, the node definition instructions may include the coordinates of the top and bottom nodes of the pier, as well as the coordinates of the nodes at the ends of the cap beam. Next, the element connection instructions define the connection relationships between nodes, forming the framework of the structure. These instructions not only describe the types of members (such as column elements, beam elements, etc.), but also specify the starting and ending nodes of the members. By parsing these instructions, the length and direction of the members can be determined, and the geometric dimensions of the members can be calculated. For example, the element connection instructions for the bridge pier may specify the connection between the top node and the bottom node of the pier, thereby determining the height of the pier. At the same time, the constraint condition setting instructions define the boundary conditions of the structure, which have a significant impact on the mechanical behavior of the structure. Boundary conditions can include fixed constraints, angular constraints, and displacement constraints, which limit the degrees of freedom of the nodes. By parsing these instructions, it can be determined which nodes are subject to which type of constraint, ensuring the stability and reliability of the structural model under actual working conditions. For example, the bottom node of the bridge pier may be set as a fixed constraint, limiting all its degrees of freedom.

[0127] In addition to generating nodes, members, and boundary conditions, material parameters and geometric dimensions need to be set according to design parameters. Material parameters include elastic modulus, Poisson's ratio, density, etc., which directly affect the mechanical behavior of the structure. Geometric dimensions include cross-sectional area, moment of inertia, etc., which determine the load-carrying capacity and deformation characteristics of the members. By parsing the relevant instructions in the modeling script, these parameters and dimensions can be accurately applied to the structural model.

[0128] Finally, by integrating the node definition instructions, element connection instructions, and constraint condition setting instructions, a complete structural model is generated. This model not only includes nodes, members, and boundary conditions, but also covers material parameters and geometric dimensions, providing a solid foundation for subsequent structural analysis and optimization.

[0129] Step S304: Calculate and apply self-weight load, T-beam constant load reaction force, and braking force according to the load application instructions in the modeling script.

[0130] It is important to note that the load application instructions provide detailed information about the type, size, and location of the loads to be applied, which is crucial for simulating actual working conditions.

[0131] Specifically, first, the dead load is applied according to the load application instructions in the modeling script. The dead load is the load generated by the weight of the structure itself, which has a significant impact on the internal force and deformation of the structure. The size and distribution of the dead load depend on the material density and geometric shape of the structure. In the modeling script, the density parameters of the material are usually specified, and through these parameters, the dead weight of each member can be calculated and uniformly distributed to the corresponding nodes. For example, for concrete piers and cap beams, the dead weight needs to be calculated according to their cross-sectional size and length, and applied as a uniformly distributed load to the structure. Second, the T-beam constant load counterforce is applied. The T-beam constant load counterforce is the load transmitted from the superstructure to the substructure, which reflects the influence of the weight of the superstructure on the substructure. In the modeling script, the T-beam constant load counterforce is usually applied as a concentrated force to specific nodes of the cap beam or pier. The size and distribution of these counterforces need to be calculated according to the design parameters and arrangement of the T-beam. For example, if the T-beam has multiple support points on the cap beam, the size of the counterforce at each support point will be distributed according to the weight of the T-beam and the location of the support point. Finally, the braking force is applied. The braking force is the horizontal force generated when the vehicle brakes, which has an important influence on the stability of the pier. In the modeling script, the braking force is usually applied as a horizontal concentrated force to the nodes at the top of the pier. The size of the braking force needs to be calculated according to the weight of the vehicle and the braking acceleration. For example, for long-span bridges on highways, the braking force may be an important design consideration that needs to be evaluated in detail according to traffic flow and vehicle type.

[0132] By accurately applying the dead load, T-beam constant load counterforce, and braking force, the structural analysis model can accurately reflect the actual stress conditions under the actual working conditions. This step not only improves the accuracy of the analysis results, but also provides reliable data support for subsequent design optimization and safety evaluation.

[0133] In step S305, the material parameters and geometric dimensions, nodes, members, and boundary conditions, as well as the dead load, T-beam constant load counterforce, and braking force are input into the structural analysis model to perform static analysis and obtain the core response values.

[0134] It should be noted that in order to perform static analysis, material parameters and geometric dimensions, nodes, members and boundary conditions, as well as self-weight load, T-beam constant load reaction force and braking force need to be input into the structural analysis model. First, according to the material parameters and geometric dimensions, the mechanical properties of each member are defined. Material parameters include elastic modulus, Poisson's ratio, density, etc., and geometric dimensions include cross-sectional area, moment of inertia, etc. These parameters directly affect the mechanical behavior of the structure, so they must be accurately input into the model. Next, nodes, members and boundary conditions are input into the model. Nodes define the geometric position of the structure, members define the connection relationship between nodes, and boundary conditions define the constraint state of the structure. These information together constitute the geometric and mechanical model of the structure. Then, the self-weight load, T-beam constant load reaction force and braking force are applied. The self-weight load is the load generated by the weight of the structure itself, the T-beam constant load reaction force is the load transmitted from the upper structure to the lower structure, and the braking force is the horizontal force generated when the vehicle brakes. Accurate application of these loads is the key to simulating actual working conditions and directly affects the response results of the structure.

[0135] After inputting all the necessary information, run the static analysis. Static analysis gradually applies loads through the solver and calculates the response of the structure at each step. The analysis results include core response values such as node displacement, internal force distribution, etc. For example, the maximum bending moment and shear force of the key section of the cap beam, the horizontal displacement at the top of the pier and the vertical reaction force at the bottom of the pier, etc. These core response values are important indicators for evaluating the safety and reliability of the structure.

[0136] Based on the core response values obtained through static analysis, engineers can determine whether the structure meets the requirements of design specifications. If some response values exceed the allowed range, design parameters may need to be adjusted or the structure may need to be redesigned.

[0137] Step S306, based on the core response values, obtain the structural performance results.

[0138] It should be noted that the structural performance results include the maximum bending moment, shear force of the key section of the cap beam, and the horizontal displacement at the top of the pier and the vertical reaction force at the bottom of the pier. First, the maximum bending moment and shear force of the key section of the cap beam are important indicators for evaluating the load-carrying capacity of the cap beam. The maximum bending moment reflects the maximum bending resistance of the cap beam under load, while the shear force reflects the shear resistance of the cap beam. By analyzing these indicators, it can be determined whether the cap beam can withstand the design load and ensure the safety of the structure. For example, if the calculated maximum bending moment exceeds the bending strength of the material, the section size may need to be increased or the material properties may need to be adjusted. Second, the horizontal displacement at the top of the pier and the vertical reaction force at the bottom of the pier are key indicators for evaluating the stability and load-carrying capacity of the pier. The horizontal displacement at the top of the pier reflects the deformation of the pier under horizontal load, while the vertical reaction force at the bottom of the pier reflects the support capacity of the pier to the load of the superstructure. By analyzing these indicators, it can be determined whether the pier can stably support the superstructure and ensure the overall stability of the bridge. For example, if the horizontal displacement at the top of the pier is too large, the stiffness of the pier may need to be increased or the boundary conditions may need to be adjusted.

[0139] Through these structural performance results, engineers can comprehensively evaluate whether the design scheme of the bridge substructure meets the engineering requirements. If some performance indicators do not meet the requirements, targeted design adjustments can be made, such as modifying the section size, adjusting the material properties, or rearranging the components, etc. These adjustments require corresponding modifications in the modeling script, and the structural response calculation needs to be performed again until all performance indicators meet the design requirements. Through this performance result-based evaluation and adjustment process, the safety, reliability, and economy of the bridge substructure design can be ensured, providing strong support for the rapid construction and safe operation of bridges.

[0140] The present embodiment realizes seamless connection from modeling script to performance evaluation by automatically building a structural analysis model and performing static calculation, quickly generates substructure parameters that meet design specifications through context learning and few-shot prompting mechanism of large language models, improves design efficiency and accuracy, and is especially suitable for complex bridge types and non-standard design scenarios, providing intelligent design support for bridge engineering.

[0141] Based on the first embodiment of the present application, the present application also provides a bridge substructure modeling scheme generation device, please refer to Figure 4 , the device comprises:

[0142] The acquisition module 10 is configured to acquire the design input parameters of the bridge superstructure.

[0143] The processing module 20 is configured to construct the design task information of the bridge substructure according to the design input parameters of the bridge superstructure.

[0144] The reasoning module 30 is configured to generate the design parameters of the bridge substructure by reasoning the design task information through the large language model.

[0145] The script generation module 40 is configured to input the design parameters into the large language model for analysis and reasoning to generate a modeling script.

[0146] The calculation module 50 is configured to perform structural response calculation through the modeling script to obtain a structural performance result.

[0147] The judgment module 60 is configured to judge according to the structural performance result to obtain a judgment result.

[0148] The result module 70 is configured to execute different modeling strategies according to the judgment result to generate a target bridge substructure modeling scheme, and is further configured to, if the judgment result is that the structural performance result does not satisfy a preset performance threshold, generate an optimization adjustment instruction, feed back the optimization adjustment instruction to the large language model to adjust the design parameters, obtain target design parameters, and simultaneously analyze the modeling script corresponding to the structural performance result, if the analysis result of the modeling script does not converge, generate a script optimization instruction and feed back to the large language model for constraint to obtain a target modeling script, if the judgment result is that the structural performance result satisfies the preset performance threshold, take the design parameters corresponding to the structural performance result as the target design parameters and take the modeling script corresponding to the structural performance result as the target modeling script, and generate the target bridge substructure modeling scheme according to the target design parameters and the target modeling script.

[0149] The bridge substructure modeling scheme generation device provided by the present application adopts the bridge substructure modeling scheme generation method in the above embodiments, and can solve the technical problem of how to realize the design parameter generation and automatic modeling of the bridge substructure in the non-standard bridge type scene, and improve the design efficiency and accuracy. Compared with the prior art, the bridge substructure modeling scheme generation device provided by the present application has the same beneficial effects as the bridge substructure modeling scheme generation method provided by the above embodiments, and other technical features in the bridge substructure modeling scheme generation device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0150] In an embodiment, the processing module 20 is further configured to obtain historical bridge superstructure input parameters, analyze the bridge superstructure input parameters to obtain a bridge deck width, a next pier bridge deck distribution span and a structure type, filter and convert from a historical design sample library according to a preset requirement based on the bridge deck width, the next pier bridge deck distribution span and the structure type to obtain an example pair, wherein the historical design sample library is established based on the historical bridge superstructure input parameters, and the example pair is a corresponding relationship between the bridge superstructure input parameters and the design parameters of the bridge substructure, extract design specification requirements related to the substructure design from a specification database, and splice the example pair, the design specification requirements and the superstructure input parameters to generate a design task prompt information of the bridge substructure.

[0151] In an embodiment, the processing module 20 is further configured to: filter, from the historical design sample library, a first sample set that is of the same structure type and has a difference in bridge deck width less than a preset width threshold; filter, from the historical design sample library, a second sample set that is of the same structure type and has a next pier bridge deck distribution span greater than a preset span threshold; and convert samples in the first sample set and the second sample set into example pairs in text form.

[0152] In an embodiment, the reasoning module 30 is further configured to: obtain historical superstructure input parameters; input the design task information into a large language model to obtain an association result of the bridge substructure design task information, design specification requirements, and the historical superstructure input parameters by using a context learning capability of the large language model; obtain a structure design rule based on the historical superstructure input parameters and a few-shot prompting mechanism of the large language model; and generate design parameters of the bridge substructure based on the association result and the structure design rule.

[0153] In an embodiment, the script generation module 40 is further configured to: calculate, according to a pier column number and size parameter in the design parameters, to obtain spatial coordinates of each pier column node; define, according to a cap beam size parameter in the design parameters, to confirm a cross-sectional geometric attribute of each segment of the cap beam; set, according to a material parameter in the design parameters, to obtain a material parameter of a structural material; generate node definition instructions and element connection instructions based on the spatial coordinates, the cross-sectional geometric attribute, and the material parameter; generate load application instructions according to a load parameter in the design parameters; generate setting instructions of a structural constraint condition according to a boundary condition parameter in the design parameters; and integrate the node definition instructions, the element connection instructions, the load application instructions, and the setting instructions to input into the large language model to generate a modeling script of the bridge substructure.

[0154] In an embodiment, the calculation module 50 is further configured to: construct a structure analysis model, initialize a model dimension and a degree of freedom; set a solver, a convergence criterion, and a loading step of the structure analysis model; obtain material parameters, geometric sizes, nodes, components, and boundary conditions according to the node definition instructions, the element connection instructions, and the setting instructions of the constraint condition in the modeling script; calculate and apply dead load, T-beam constant load reaction force, and braking force according to the load application instructions in the modeling script; input the material parameters and the geometric sizes, the nodes, the components, and the boundary conditions, and the dead load, the T-beam constant load reaction force, and the braking force into the structure analysis model to perform static analysis, to obtain core response values; and obtain a structure performance result based on the core response values, the result including a maximum bending moment and a shear force of a key section of the cap beam, and a pier top horizontal displacement and a pier bottom vertical reaction force.

[0155] In an embodiment, the result module 70 is further configured to generate an initial modeling scheme of the bridge substructure according to the target design parameters and the target modeling script, the initial modeling scheme including node arrangement, member definition, load application, and boundary condition setting; associate the modeling scheme with an actual construction scenario to generate a scenario association report, wherein the actual construction scenario includes geological conditions, climate conditions, and traffic conditions of a construction site; generate a cost estimation report according to the target design parameters and the target modeling script in combination with material market prices and construction costs; perform a safety evaluation on the modeling scheme to generate a safety evaluation report; evaluate an impact of the modeling scheme on a surrounding environment to generate an environmental impact evaluation report; generate a comprehensive evaluation report by synthesizing the scenario association report, the cost estimation report, the safety evaluation report, and the environmental impact evaluation report; perform a multi-dimensional score calculation on the comprehensive evaluation report to obtain a comprehensive score; and when the comprehensive score exceeds a preset score threshold, take the initial modeling scheme as a target bridge substructure modeling scheme.

[0156] The present application provides a bridge substructure modeling scheme generation device, which comprises at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the bridge substructure modeling scheme generation method in the above-mentioned embodiment one.

[0157] Reference will now be made to the drawings, in which Figure 5 which shows a structural schematic diagram of a bridge substructure modeling scheme generation device suitable for use to implement embodiments of the present application. The bridge substructure modeling scheme generation device in the embodiments of the present application can include, but is not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast acquirers, PDAs (Personal Digital Assistant), PADs (Portable Application Description), PMPs (Portable MediaPlayer), vehicle-mounted terminals (for example, vehicle-mounted navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. Figure 5 The bridge substructure modeling scheme generation device shown is merely an example, and should not bring any limitation to the functions and use range of the embodiments of the present application.

[0158] As Figure 5As shown, the bridge substructure modeling scheme generation apparatus can include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. In the RAM 1004, various programs and data required for the operation of the bridge substructure modeling scheme generation apparatus are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. In general, the following can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; the storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the bridge substructure modeling scheme generation apparatus to communicate wirelessly or by wire with other devices to exchange data. Although a bridge substructure modeling scheme generation apparatus having various components is shown in the figure, it should be understood that all of the shown components are not required to be implemented or possessed. More or less can be alternatively implemented or possessed.

[0159] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments of the present disclosure are performed.

[0160] The bridge substructure modeling scheme generation device provided by the application adopts the bridge substructure modeling scheme generation method in the above embodiment, and can solve the technical problem of how to realize the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scene, and improve the design efficiency and accuracy. Compared with the prior art, the bridge substructure modeling scheme generation device provided by the application has the same beneficial effects as the bridge substructure modeling scheme generation method provided by the above embodiment, and other technical features in the bridge substructure modeling scheme generation device are the same as the features disclosed in the previous embodiment method, and will not be repeated here.

[0161] It should be understood that parts of the present application can be realized by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0162] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0163] The present application provides a computer readable storage medium having computer readable program instructions (i.e. computer programs) stored thereon, the computer readable program instructions being used to execute the bridge substructure modeling scheme generation method in the above embodiment.

[0164] The computer readable storage medium provided in the present application may, for example, be a U disk, but is not limited to an electric, magnetic, optical, electromagnetic, infrared, or semiconductor, or device, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electric connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read only memory (ROM), an erasable programmable read only memory (EPROM or flash memory), an optical fiber, a portable compact disk read only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present embodiment, the computer readable storage medium can be any tangible storage medium containing or storing a program that can be executed by or used in conjunction with an instruction execution device. The program code contained on the computer readable storage medium can be transmitted in any suitable storage medium, including but not limited to an electric wire, an optical cable, an RF (Radio Frequency), and the like, or any suitable combination of the above.

[0165] The above computer readable storage medium can be contained in the bridge substructure modeling scheme generation device; or can exist separately without being assembled into the bridge substructure modeling scheme generation device.

[0166] The above computer readable storage medium carries one or more programs, which, when executed by the bridge substructure modeling scheme generation device, enable the bridge substructure modeling scheme generation device to be written in one or more programming languages or combinations thereof for executing computer program codes for performing the operations of the present application, the programming languages including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" language or similar programming languages. The program code can be executed entirely on a user computer, partially on a user computer, as an independent software package, partially on a user computer and partially on a remote computer, or entirely on a remote computer or server. In the case involving a remote computer, the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, through the Internet using an Internet service provider).

[0167] The flow and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of various embodiments of the methods and computer program products according to the present application. In this regard, each block in the flow and block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may be executed in the reverse order, depending on the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by special purpose hardware-based systems that perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.

[0168] The modules involved in the embodiments of the present application can be implemented in software or in hardware. In some cases, the names of the modules do not constitute a limitation on the modules themselves.

[0169] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., a computer program) for executing the bridge substructure modeling scheme generation method described above, and can solve the technical problem of how to realize the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scenario, and improve the design efficiency and accuracy. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the bridge substructure modeling scheme generation method provided by the above embodiments, and will not be described here.

[0170] The present application also provides a computer program product comprising a computer program, which, when executed by a processor, implements the steps of the bridge substructure modeling scheme generation method as described above.

[0171] The computer program product provided by the present application can solve the technical problem of how to realize the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scenario, and improve the design efficiency and accuracy. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the bridge substructure modeling scheme generation method provided by the above embodiments, and will not be described here.

[0172] The above merely describes some embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structural transformation, direct / indirect application in other related technical fields, or the like, which is made based on the technical concept of the present application and the content of the specification and drawings, is included in the patent protection scope of the present application.

Claims

1. A method for generating a modeling scheme for the substructure of a bridge, characterized in that, include: Obtain the design input parameters for the bridge superstructure; The design task information for the bridge substructure is constructed based on the design input parameters of the bridge superstructure. The design task information is reasoned through a large language model to generate design parameters for the bridge substructure. The design parameters are input into the large language model for analysis and reasoning to generate a modeling script. The structural response is calculated using the modeling script to obtain the structural performance results; Based on the structural performance results, a judgment is obtained; Based on the judgment results, different modeling strategies are executed to generate a modeling scheme for the substructure of the target bridge. The judgment results include structural performance results that do not meet the preset performance threshold and structural performance results that meet the preset performance threshold. The step of executing different modeling strategies based on the judgment result to generate a modeling scheme for the substructure of the target bridge includes: If the judgment result is that the structural performance result does not meet the preset performance threshold, an optimization adjustment instruction is generated and fed back to the large language model to adjust the design parameters to obtain the target design parameters. Simultaneously, the modeling script corresponding to the structural performance results is analyzed. If the analysis results of the modeling script do not converge, script optimization instructions are generated and fed back to the large language model for constraint, thus obtaining the target modeling script. If the judgment result is that the structural performance result meets the preset performance threshold, then the design parameters corresponding to the structural performance result are taken as the target design parameters and the modeling script corresponding to the structural performance result is taken as the target modeling script; Generate a modeling scheme for the substructure of the target bridge based on the target design parameters and the target modeling script; The step of constructing design task information for the bridge substructure based on the design input parameters of the bridge superstructure includes: Obtain input parameters for the superstructure of historical bridges; By analyzing the input parameters of the bridge superstructure, the bridge deck width, the span of the next pier, and the structural type are obtained. Based on the bridge deck width, the span of the next pier bridge deck distribution, and the structural type, the historical design sample library is filtered and transformed according to preset requirements to obtain example pairs. The historical design sample library is established based on the input parameters of the superstructure of the historical bridge, and the example pairs are the correspondence between the input parameters of the superstructure of the bridge and the design parameters of the substructure of the bridge. Extract design specification requirements related to the substructure design from the specification database; The example pairs, the design specifications, and the superstructure input parameters are combined to generate design task information for the bridge substructure. The step of filtering and converting sample pairs from the historical design sample library according to preset requirements based on the bridge deck width, the span of the next pier bridge deck distribution, and the structural type includes: Select a first set of samples from the historical design sample library that are of the same structure type and whose difference from the bridge deck width is less than a preset width threshold; Select a second set of samples from the historical design sample library that are of the same structure type as the above-mentioned structure and whose span of the next pier bridge deck distribution is greater than a preset span threshold. Convert the samples in the first sample set and the second sample set into text-based example pairs.

2. The method as described in claim 1, characterized in that, The step of reasoning about the design task information using a large language model to generate design parameters for the bridge substructure includes: Obtain the historical upper structure input parameters; The design task information is input into a large language model, and the context learning capability of the large language model is used to obtain the correlation results of the bridge substructure design task information, design specification requirements, and historical substructure input parameters. Based on the historical upper structure input parameters, the structural design rules are obtained through the few-sample prompting mechanism of the large language model; Based on the correlation results and the structural design rules, the design parameters of the bridge substructure are generated through reasoning.

3. The method as described in claim 1, characterized in that, The step of inputting the design parameters into the large language model for analysis and reasoning to generate a modeling script includes: The spatial coordinates of each pier node are calculated based on the number and size parameters of the piers in the design parameters. Define the cap beam dimensions based on the design parameters and confirm the cross-sectional geometric properties of each segment of the cap beam. The material parameters of the structural material are obtained by setting the material parameters in the design parameters. Based on the spatial coordinates, the cross-sectional geometric properties, and the material parameters, node definition instructions and element connection instructions are generated. Based on the load parameters in the design parameters, generate load application instructions; Based on the boundary condition parameters in the design parameters, generate instructions for setting structural constraints; The node definition instructions, unit connection instructions, load application instructions, and setting instructions are integrated and input into the large language model for reasoning to generate a modeling script for the bridge substructure.

4. The method as described in claim 1, characterized in that, The step of calculating the structural response using the modeling script to obtain the structural performance results includes: Construct a structural analysis model and initialize the model dimensions and degrees of freedom; Set the solver, convergence criterion, and loading step size for the structural analysis model; Material parameters, geometric dimensions, nodes, components, and boundary conditions are obtained based on the node definition instructions, element connection instructions, and constraint setting instructions in the modeling script. Calculations are performed based on the load application instructions in the modeling script, applying self-weight load, T-beam dead load reaction force, and braking force. Static analysis is performed on the structural analysis model based on the material parameters and geometric dimensions, the nodes, components and boundary conditions, as well as the self-weight load, the dead load reaction force of the T-beam and the braking force, to obtain the core response value. Based on the core response values, structural performance results are obtained, including the maximum bending moment and shear force of the key sections of the cap beam, as well as the horizontal displacement at the top of the pier and the vertical reaction force at the bottom of the pier.

5. The method as described in claim 1, characterized in that, The step of generating a modeling scheme for the substructure of the target bridge based on the target design parameters and the target modeling script includes: Based on the target design parameters and the target modeling script, an initial modeling scheme for the bridge substructure is generated, which includes node arrangement, component definition, load application, and boundary condition setting. The modeling scheme is associated with the actual construction scenario to generate a scenario association report, wherein the actual construction scenario includes the geological conditions, climate conditions and traffic conditions of the construction site; Based on the target design parameters and the target modeling script, combined with material market prices and construction costs, a cost estimation report is generated. A security assessment is performed on the modeling scheme, and a security assessment report is generated. Assess the impact of the modeling scheme on the surrounding environment and generate an environmental impact assessment report; A comprehensive assessment report is generated by combining the scenario-related reports, cost estimation reports, safety assessment reports, and environmental impact assessment reports. The comprehensive evaluation report is scored using multiple dimensions to obtain a comprehensive score; When the overall score exceeds a preset score threshold, the initial modeling scheme will be used as the target bridge substructure modeling scheme.

6. A device for generating modeling schemes for bridge substructures, characterized in that, The device includes: The acquisition module is used to acquire the design input parameters for the bridge superstructure. The processing module is used to construct design task information for the bridge substructure based on the design input parameters of the bridge superstructure; it is also used to obtain historical bridge superstructure input parameters; analyze the bridge superstructure input parameters to obtain the bridge deck width, the span of the next pier deck distribution, and the structural type; and based on the bridge deck width, the span of the next pier deck distribution, and the structural type, filter and transform the historical design sample library according to preset requirements to obtain example pairs, wherein the historical design sample library is established based on the historical bridge superstructure input parameters, and the example pairs are pairs of bridge superstructure input parameters and bridge substructure design parameters. The system is configured to: extract design specification requirements related to the substructure design from the specification database; concatenate the example pairs, the design specification requirements, and the superstructure input parameters to generate design task information for the bridge substructure; further, select a first sample set from a historical design sample library that is the same type as the structure and whose difference from the bridge deck width is less than a preset width threshold; select a second sample set from the historical design sample library that is the same type as the structure and whose next pier bridge deck span is greater than a preset span threshold; and convert the samples in the first and second sample sets into text-based example pairs. The reasoning module is used to reason about the design task information through a large language model and generate design parameters for the bridge substructure. The script generation module is used to input the design parameters into the large language model for analysis and reasoning, and generate a modeling script. The calculation module is used to perform structural response calculations using the modeling script to obtain structural performance results; The judgment module is used to make a judgment based on the structural performance results and obtain a judgment result; The result module is used to execute different modeling strategies based on the judgment result to generate a modeling scheme for the target bridge substructure; it is also used to generate an optimization adjustment instruction if the judgment result indicates that the structural performance result does not meet a preset performance threshold, and feed the optimization adjustment instruction back to the large language model to adjust the design parameters to obtain the target design parameters; at the same time, it analyzes the modeling script corresponding to the structural performance result, and if the analysis result of the modeling script does not converge, it generates a script optimization instruction and feeds it back to the large language model for constraint to obtain the target modeling script; if the judgment result indicates that the structural performance result meets the preset performance threshold, it uses the design parameters corresponding to the structural performance result as the target design parameters and the modeling script corresponding to the structural performance result as the target modeling script; and generates a modeling scheme for the target bridge substructure based on the target design parameters and the target modeling script.

7. A device for generating modeling schemes for bridge substructures, characterized in that, The device includes: a memory, a processor, and a bridge substructure modeling scheme generation program stored in the memory and running on the processor, the bridge substructure modeling scheme generation program being configured to implement the steps of the bridge substructure modeling scheme generation method as described in any one of claims 1-5.

8. A storage medium, characterized in that, The storage medium stores a bridge substructure modeling scheme generation program, which, when executed by a processor, implements the steps of the bridge substructure modeling scheme generation method as described in any one of claims 1-5.

Images