Building structure automatic design method and system based on artificial intelligence

Abstract

The invention relates to the technical field of data processing, and discloses a building structure automatic design method and system based on artificial intelligence. The method comprises the steps that a collaborative building information model is constructed by collecting building design task data, load analysis and rigidity parameter calculation are carried out, a Timoshenko beam model is constructed, a structural arrangement prediction result is generated in combination with graph network optimization, a parameterized modeling program is input to generate a building structure model, and multi-working-condition analysis is completed to obtain a structural design scheme. Design documents and construction information are automatically generated, and a parameter and analysis result corresponding relation database is established. By integrating modeling, structural analysis and design document generation, the problems of intelligent collaboration and efficient output in the whole process of structural design are effectively solved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an artificial intelligence-based automated design method and system for building structures. Background Art

[0002] In existing building structure design practices, structural engineers typically rely on traditional CAD drawing tools, structural analysis software (such as SAP2000 and PKPM), and manual experience to design and verify building structural schemes. This process often involves extracting building component information from 2D design drawings, manually building a structural model, inputting load conditions for internal force analysis, and finally completing component selection and node connection design. Although BIM (Building Information Modeling) has been applied in some projects, there is still a disconnect between structural mechanics analysis, load calculation, stiffness control, and construction drawing generation. In particular, there is a lack of effective linkage with structural dynamic characteristics, automatic component classification, parametric modeling, and construction information generation. As a result, design efficiency and accuracy are still largely limited by manual experience and data incompatibilities between various software platforms.

[0003] However, the core problem in the existing structural design process is the lack of fully automated modeling and analysis methods. Specifically, the spatial layout, load combinations, and node connection forms of building components rely on manual input, which is prone to faults and redundancy in the data transmission process, and there is no systematic analysis mechanism for the distribution of structural stiffness and the response of component connections. Second, during the process of structural layout optimization, mode analysis, and construction drawing generation, there is no established closed-loop system that can automatically achieve the transition from prediction to modeling, from design to drawing through algorithms. This results in a large number of information silos between structural models and design documents, making rapid iteration and efficient delivery impossible. This is especially true in large-scale steel structure projects such as shipyard bent structure design, where there are many components, complex nodes, and variable working conditions. Existing methods struggle to meet the requirements of high efficiency, high precision, and full process connectivity. Summary of the Invention

[0004] This application provides an artificial intelligence-based automated design method and system for building structures, which effectively solves the problem of intelligent collaboration and efficient output in the entire structural design process through integrated modeling, structural analysis and design document generation.

[0005] In the first aspect, the present application provides an artificial intelligence-based automated design method for building structures, which includes: collecting data on building design tasks to obtain a building information model containing collaborative information; performing load analysis and stiffness characteristic parameter calculation on the building information model, and obtaining a structural stiffness distribution scheme by determining the bending stiffness coefficient and the shear stiffness ratio; based on the structural stiffness distribution scheme, constructing a lumped Timoshenko beam simplified model and integrating a graph network optimization algorithm, and obtaining a structural layout prediction result through component relationship analysis and mode matching calculation; inputting the structural layout prediction result into a parametric modeling program for component generation to obtain a building structure model; performing multi-working condition analysis and calculation on the building structure model to obtain a structural design scheme; generating design documents and construction information based on the structural design scheme, and establishing a database of correspondence between design parameters and analysis results to obtain a building structure design scheme.

[0006] In a second aspect, the present application provides an artificial intelligence-based automated building structure design system, the artificial intelligence-based automated building structure design system comprising: The acquisition module is used to collect data for architectural design tasks and obtain a building information model containing collaborative information; a calculation module, configured to perform load analysis and stiffness characteristic parameter calculation on the building information model, and obtain a structural stiffness distribution scheme by determining a bending stiffness coefficient and a shear stiffness ratio; A construction module is used to construct a simplified model of a lumped Timoshenko beam based on the structural stiffness distribution scheme and integrate a graph network optimization algorithm to obtain a structural layout prediction result through component relationship analysis and mode shape matching calculation; An input module, configured to input the structural arrangement prediction results into a parametric modeling program to generate components and obtain a building structure model; An analysis module is used to perform multi-condition analysis and calculation on the building structure model to obtain a structural design solution; The generation module is used to generate design documents and construction information based on the structural design scheme, and establish a database of corresponding relationships between design parameters and analysis results to obtain a building structure design scheme.

[0007] In a third aspect, an artificial intelligence-based automated building structure design device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the artificial intelligence-based automated building structure design device executes the above-mentioned artificial intelligence-based automated building structure design method.

[0008] In a fourth aspect, a computer-readable storage medium is provided, wherein instructions are stored in the computer-readable storage medium, which, when executed on a computer, enables the computer to execute the above-mentioned method for automated design of building structures based on artificial intelligence.

[0009] In the technical solution provided in this application, data collection is performed on architectural design tasks to obtain a building information model containing collaborative information. This allows information such as building geometry, equipment layout, material parameters, and load application locations to be stored and called up in a unified data format, effectively solving the problems of multi-disciplinary data disconnection and repeated modeling in the traditional design process. By performing load analysis and stiffness characteristic parameter calculations on this building information model, and further determining the bending stiffness coefficient and shear stiffness ratio, a stiffness response distribution data set within the entire structural domain is constructed, improving the overall force balance of the structure and the ability to identify local responses. On this basis, a simplified model of a lumped Timoshenko beam is constructed using the structural stiffness distribution scheme, and a graph network optimization algorithm is integrated to map the original spatial component relationships into a graph network with physical properties. Through component relationship analysis and mode matching calculations, the prediction, deduction, and adjustment optimization of the structural layout are achieved, significantly improving the rationality of the structural layout and its seismic performance. The predicted structural layout results are then input into a parametric modeling program for component generation. This intelligently determines component classification based on the type of force, automatically selects standard or customized sections, and quickly generates calculable and plottable three-dimensional component models, significantly reducing modeling workload and improving consistency. Next, by performing multi-condition analysis and calculations on the generated building structure model, a comprehensive assessment of structural stiffness, displacement, and component internal forces is completed under multiple load combination scenarios. Component size and connection methods are adjusted based on the response results to ensure structural safety and economy. Finally, the structural design solution automatically generates design documents and construction information, establishes a database of corresponding relationships between design parameters and analysis results, completes structured archiving and data integration of design results, and implements a closed-loop process from inputting architectural design tasks to outputting complete structural design solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 This is a schematic diagram of an embodiment of the method for automated design of building structures based on artificial intelligence in an embodiment of the present application; Figure 2 This is a schematic diagram of an embodiment of an artificial intelligence-based automated building structure design system in an embodiment of the present application; Figure 3 It is a schematic block diagram of the structure of an artificial intelligence-based automated building structure design device in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] The embodiments of the present application provide a method and system for automated design of building structures based on artificial intelligence. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices.

[0013] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of the method for automated design of building structures based on artificial intelligence includes: Step S101: collect data on the architectural design task to obtain a building information model containing collaborative information; Step S102: performing load analysis and stiffness characteristic parameter calculation on the building information model, and obtaining a structural stiffness distribution scheme by determining the bending stiffness coefficient and the shear stiffness ratio; Step S103: Based on the structural stiffness distribution scheme, a simplified model of the lumped Timoshenko beam is constructed and integrated with a graph network optimization algorithm to obtain a structural layout prediction result through component relationship analysis and mode shape matching calculation; Step S104: input the structural layout prediction results into a parametric modeling program to generate components and obtain a building structure model; Step S105: Perform multi-condition analysis and calculation on the building structure model to obtain a structural design solution; Step S106: Generate design documents and construction information based on the structural design plan, and establish a database of corresponding relationships between design parameters and analysis results to obtain a building structure design plan.

[0014] It is understandable that the execution subject of this application can be an artificial intelligence-based building structure automation design system, or a terminal or server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0015] Specifically, data collection uses the architectural design task as input, employing multi-source information fusion to construct a Building Information Model (BIM) containing collaborative information. Basic parameters such as the number of floors, span, functional use, and component material type are extracted from the design brief. Project templates are retrieved from the structural design database to generate initial values ​​for these parameters. Next, site survey data is generated using 3D laser scanning technology to generate a point cloud dataset. The spatial coordinates (x, y, z) of each point in the point cloud are categorized as either main structure or non-structural components using a boundary recognition algorithm. A semantic segmentation model is then used to classify components such as equipment, walls, and floor slabs within the point cloud, mapping them to corresponding component information based on the classification labels. Next, specialized drawings (such as plumbing, electrical, and HVAC diagrams) are imported, and image recognition and annotation parsing algorithms are used to extract equipment layout information and technical parameters, mapping the spatial layout to load application points. All of this information is encoded and integrated into a multidimensional BIM data structure containing fields such as spatial geometry, material properties, load location, and load direction, forming a complete building information model containing collaborative information.

[0016] The load analysis and stiffness parameter calculation phases first require analyzing the load conditions of each component in the building information model. Specifically, the model's structural system is identified, extracting the node connection information and relative layout of each component to determine whether it is a beam, column, or wall. The corresponding load types are then assigned, such as dead loads from the structure's deadweight, live loads from functional requirements, wind loads based on ground roughness and building height classification, and seismic loads based on response spectrum data set according to regional standards. These loads are combined according to their direction (X / Y / Z axes) and type, and a load case matrix is ​​formed in accordance with the "Code for Loads on Building Structures." Based on this, the stiffness analysis is conducted. For each component, the cross-sectional data is used to calculate its deformation resistance in the bending (M) and shear (V) directions, based on its dimensions, material, boundary conditions, and connection method. The resulting bending stiffness (EI) and shear stiffness (GA) values ​​are then derived. These stiffness values ​​are mapped according to their spatial distribution into a structural stiffness distribution map. This map is a three-dimensional array, with each cell containing the EI and GA of a spatial unit, forming a structural stiffness distribution scheme.

[0017] The above structural stiffness data is input into a simplified Timoshenko beam model. Timoshenko beam theory considers shear deformation and rotational inertia, making it more suitable for deep beams or areas with sudden stiffness changes than conventional Euler-Bernoulli beams. First, based on the stiffness distribution, the structure is discretized into multiple lumped mass and stiffness elements to form a Timoshenko beam element mesh. The mass and stiffness matrices of each element are set. Subsequently, these structural elements are used as vertices in a graph neural network (GNN), and the connections between each two components are used as edges to construct a structural topology graph. Edge weights are assigned to each edge, representing the stiffness transfer capacity between the two components. The attribute vector of each node is initialized. The message passing mechanism of the GNN is used to transfer attribute information between adjacent nodes, updating node characteristics. Modal analysis is then performed to solve the nodal vibration modes. Spectral clustering is used to group structural elements with similar or coupled frequencies into a module. The structural layout is then optimized to ensure that the layout prediction results are consistent with the dynamic characteristics. The final spatial coordinates of each node and the component connection logic are output, forming the structural layout prediction result.

[0018] The parametric modeling program receives the structural layout prediction results and first divides the components into axial-dominated, moment-dominated, and mixed-force components based on the predicted force information. The standard component database is queried for each type of component. If the match fails, its section parameters are calculated according to the force data to generate a customized section. Next, the component geometry data is modeled. For example, moment-dominated components are generated as I-type or box-type sections, column components are modeled as H-type or circular tube sections, and wall components use plate sections. The node connection method is determined based on the component type and intersection angle to determine whether it is a rigid connection, hinged connection, or flexible transition. Finally, the components and nodes are assembled as a whole through geometric Boolean operations to obtain a three-dimensional expression of the building structure model.

[0019] During the multi-condition analysis phase, a load combination list is constructed, including constant loads, live loads, wind loads, seismic loads and their combinations. For example, "G+0.7Q+0.6W" represents a combined condition of constant loads, live loads and wind loads. The internal forces of the components are analyzed through the static and nonlinear analysis modules to determine whether each component meets the design requirements. The internal force data is used as input parameters for the cross-section design optimization module to calculate the required minimum cross-section inertia moment, minimum web thickness, and minimum cross-sectional area. For the node connection part, finite element analysis is performed based on the bending moment, shear force, and axial force conditions of the node to determine the connection method and plate thickness. The deformation data of each component is compared with the design limit. When the deformation exceeds the limit, the cross-section optimization or lateral support design process is automatically triggered to finally generate an optimized structural design data set. The above design data is input into the document generation module. The design description document summarizes load data, internal force analysis results, component design basis, and other content. Construction drawings are derived through parametric modeling, including floor plans, node details, and component lists. Material statistics automatically summarize the dimensions and quantities of various component types. The construction sequence is automatically generated based on the component connection logic. All generated parameters are mapped one-to-one with the analysis results, creating an indexed design parameter database for data traceability and updating throughout the entire lifecycle.

[0020] In the embodiment of the present application, data collection is performed on the architectural design task to obtain a building information model containing collaborative information, so that information such as building geometry, equipment layout, material parameters and load action location can be stored and called in a unified data format, effectively solving the problem of multi-disciplinary data disconnection and repeated modeling in the traditional design process; and by performing load analysis and stiffness characteristic parameter calculation on the building information model, and further determining the bending stiffness coefficient and shear stiffness ratio, a stiffness response distribution data set within the entire structural domain is constructed, improving the overall force balance of the structure and the ability to identify local responses. On this basis, a simplified model of a lumped Timoshenko beam is constructed using the structural stiffness distribution scheme, and a graph network optimization algorithm is integrated to map the original spatial component relationship into a graph network with physical properties. Through component relationship analysis and mode matching calculation, the prediction, deduction and adjustment optimization of the structural layout are realized, significantly improving the rationality of the structural layout and seismic performance. The predicted structural layout results are then input into a parametric modeling program for component generation. This intelligently determines component classification based on the type of force, automatically selects standard or customized sections, and quickly generates calculable and plottable three-dimensional component models, significantly reducing modeling workload and improving consistency. Next, by performing multi-condition analysis and calculations on the generated building structure model, a comprehensive assessment of structural stiffness, displacement, and component internal forces is completed under multiple load combination scenarios. Component size and connection methods are adjusted based on the response results to ensure structural safety and economy. Finally, the structural design solution automatically generates design documents and construction information, establishes a database of corresponding relationships between design parameters and analysis results, completes structured archiving and data integration of design results, and implements a closed-loop process from inputting architectural design tasks to outputting complete structural design solutions.

[0021] In a specific embodiment, the process of executing step S101 may specifically include the following steps: Conduct project demand analysis on architectural design tasks to obtain basic architectural design parameters; Conduct on-site condition surveys of basic building design parameters, and obtain building space point cloud data by measuring site topography and surrounding environmental constraints; Perform feature extraction and contour recognition on the building space point cloud data, and obtain the building space geometric model by separating the spatial information of the building's main structure and non-structural components; Perform professional data association processing on the building space geometric model, import the layout drawings and technical parameter tables of various professional equipment, and obtain multi-professional collaborative data; Load identification and classification are performed based on multi-disciplinary collaborative data. By calculating the weight of each device and determining its position in the structure, a structural load distribution plan is obtained. The structural load distribution scheme is integrated to obtain a building information model containing collaborative information.

[0022] Specifically, in the project demand analysis process, the input is the design task book, the owner's requirement document and the functional planning map. These text data are parsed by natural language processing tools to extract key structural design parameters, such as load level, floor height, building use, fire resistance level, seismic fortification intensity, etc. These parameters are structured and encoded and stored in the attribute table, and matched with the reference projects set in the project template library to fill in the missing fields. For example, when the task book specifies "a single-story industrial plant with a frame structure, a span of 30 meters, a height of 10 meters, and a crane in operation", the model will automatically match the typical layout of the frame structure, extract the standard beam-column spacing, rigid connection node form, etc. from the existing data, and assign the initial value to the current project.

[0023] A site survey based on basic parameters is conducted to collect three-dimensional topographic data and environmental information. The operator uses a laser scanner to perform an area scan of the building site. Each point in the generated point cloud contains spatial coordinates (x, y, z), reflection intensity i, and scanning angle θ. The data is formatted as a standard LAS file. Point cloud preprocessing includes denoising (filtering out floating points), registration (unifying the coordinate system across different scanned surfaces), and downsampling. Spatial clustering algorithms, such as the density-based DBSCAN method, are then introduced to perform a preliminary zoning of terrain variations, automatically classifying ground, high platforms, and sunken areas. The output data is a clearly structured surface geometry point cloud, providing the foundation for subsequent contour extraction. Contour recognition processing on the point cloud data aims to effectively separate the main building structure from non-structural components in the point cloud and construct an accurate building spatial geometry model. The processing workflow begins with a normal-based surface fitting algorithm to detect large planes, vertical surfaces, or continuous edges in the point cloud. These feature sets are then labeled based on typical component size ranges (e.g., column diameters of at least 300 mm and beam spans of at least 2 meters). The CNN convolutional network classifies and recognizes the slice point cloud images, achieving a preliminary segmentation of beams, columns, slabs, and walls. Subsequently, geometric reconstruction is performed based on information such as the center of gravity distribution and component spacing to form a complete spatial geometric model, including the 3D bounding box, centerline, and cross-sectional orientation of each component.

[0024] After the spatial geometric model is constructed, drawings and technical parameter tables from various disciplines are imported to achieve multi-disciplinary data collaboration. The input data consists of CAD or BIM format drawings (such as DWG, IFC, etc.) and equipment parameter tables (such as Excel spreadsheets). The equipment layout information in the drawings is parsed using block recognition and attribute extraction techniques (such as AutoLISP scripts combined with OCR extraction). The coordinate position, dimensions, and number of each component in the drawings, such as ducts, bridges, and transformer boxes, are converted into structured entries in the database and matched with the spatial coordinates in the geometric model. This information is integrated through spatial similarity (e.g., component center of gravity distance <0.1m), forming a multi-disciplinary collaborative information structure. Load identification and classification based on this collaborative data requires extracting information such as component mass, support method, and load direction from the technical parameters. For example, if the equipment parameter table indicates that a transformer weighs 5,000 kg and is supported below a main beam, the main beam number is compared with the spatial geometric model to confirm the load application point and transmission path. All such point loads are classified as live loads. For dead loads, such as the deadweight of walls and floor structures, the volume product is calculated directly by multiplying the model's cross-sectional dimensions by the material density. Wind and seismic loads are generated using external load models based on height, openness, and the structure's natural vibration period, in accordance with regulatory standards. All loads are uniformly converted into a structured data format, with each load item containing fields such as the load value (N or kN / m²), applied location (coordinates), action direction (vector), and the affected component number.

[0025] After the load data is prepared, a structural load distribution plan is formed through load partitioning and superposition. This process tracks the load paths based on component connections, inferring the nodes and beams and columns affected by each point load. Multiple load types are then superimposed to form a set of structural unit load vectors. This set is ultimately mapped back to the building geometry, achieving a closed-loop conversion from spatial geometry to structural loads.

[0026] In a specific embodiment, the process of executing step S102 may specifically include the following steps: Identify the structural type of the building information model and obtain the building structure system data by analyzing the building plan layout and vertical component distribution characteristics; Based on the building structure system data, load sub-item calculation is performed. By separating the dead load, live load, wind load and earthquake load and determining their action directions, multi-type load action data is obtained. Perform load combination processing on multi-type load action data, and obtain critical load condition data by combining the load effects under different working conditions according to the requirements of design specifications; The structural bending stiffness is calculated based on the critical load condition data. The bending stiffness coefficient is obtained by analyzing the resistance of the structure under bending deformation. Perform shear deformation analysis on the building information model and obtain the shear stiffness ratio by calculating the shear deformation characteristics of the structure under horizontal force. The stiffness distribution is optimized based on the bending stiffness coefficient and the shear stiffness ratio. The structural stiffness distribution scheme is obtained by balancing the stiffness distribution of each region and considering the special treatment of the stiffness mutation area.

[0027] Specifically, the process of structural type identification for a building information model begins with the geometric information of each component in the building model, including its position, length, height, connection relationships, and component type. This data is organized in a structured table format. The structural system type is determined by analyzing the relative position, orientation, repetitiveness, and closed rigid unit formation of beams and columns in the building plan. This is combined with the vertical continuity and connection node density of vertical components to determine the structural system type. If the beams connecting all columns form a regular horizontal arrangement, the vertical columns are continuous from the foundation to the roof, and no shear wall elements are present, the system is identified as a steel frame structure. This identification is achieved through rule matching, where the structural form in the model is sequentially verified according to a series of compliance judgment rules and a database query is performed to determine whether it corresponds to a defined structural system type. Once structural system identification is complete, the load sub-calculation phase begins immediately. At this stage, each component in the building model is processed based on its geometric dimensions, material properties, and positional relationships. The constant load is obtained by multiplying the volume of the component itself by the material density, including roof steel plates, self-weight components, etc., and the live load is determined by the purpose of the building, such as the periodic load generated by the operation of equipment in industrial plants. The wind load is determined by the roof elevation, the windward area of ​​the structure and the geographical parameters. However, in this technical solution, the standard wind pressure values ​​matched in the database are uniformly adopted, and the product is calculated by combining the positive projection area of ​​the structure to obtain the force. The seismic load is based on the building height and structural stiffness, and the period segment of the structure is found using the existing period correspondence table, and then the equivalent seismic force is obtained according to the standard earthquake response spectrum. The focus of this process is to find a specific component in the original model as the object of action for each load, and to clearly mark the direction information of the force in the spatial coordinate system of the model.

[0028] After all loads are independently calculated, the system enters the load combination processing process. This process automatically generates different design conditions through a preset load combination matrix. For example, the basic condition combines dead loads and live loads, the intermediate condition combines dead loads and wind loads, and the extreme condition superimposes dead loads, live loads, wind loads and seismic forces. Each combination corresponds to a specific number and action mechanism. The structural model performs a force analysis under each combination and extracts the maximum internal force response of each component under each condition. Subsequently, the program will filter out the condition corresponding to each beam and column at its maximum response from all condition combinations as the control load condition for the component. This screening is achieved by comparing the response values ​​under different conditions row by row, and recording the critical condition data in the component attribute table.

[0029] Based on these critical load condition data, the system calculates the structural bending stiffness by first extracting the corresponding component's length, maximum bending value, and deformation under these conditions. Using static analysis tools to calculate the deformation, the component's ability to resist bending deformation can be determined. The greater this resistance, the greater the component's stiffness and the higher the structural stability. The bending stiffness results for all components are summarized to form a horizontal stiffness distribution map for the entire building, allowing observation of areas with significant stiffness deficiency or excessive stiffness concentration.

[0030] When a building experiences horizontal displacement due to earthquakes or wind loads, components experience shear deformation. Therefore, shear deformation analysis is required for the building information model. This process analyzes the horizontal displacement of vertical component nodes and, combined with the component's own stiffness and connection structure, determines its horizontal shear deformation capacity. The ratio of each component's horizontal deformation capacity to its bearing capacity is calculated and recorded to describe its shear stiffness ratio. A ratio that is too high or too low is detrimental to the optimal distribution of the overall structural stiffness.

[0031] After obtaining the bending and shear stiffness of all components, the overall structure needs to be optimized for stiffness distribution. This process uses spatial grids as units and adjusts the stiffness values ​​in each small area. The goal is to make the stiffness changes between adjacent areas smooth and avoid stress concentration problems caused by local stiffness mutations. In actual processing, the area with the largest stiffness difference will be identified first, and then the size, cross-sectional form or material grade of the components in this area will be adjusted to maintain continuity with the surrounding area within a reasonable range. All optimization results will be synchronously updated back to the building model to form a structural stiffness distribution plan with high data consistency and engineering applicability.

[0032] In a specific embodiment, the process of executing step S103 may specifically include the following steps: The structural stiffness distribution scheme is simplified, and the complex building structure is equivalent to a Timoshenko beam element with distributed mass and stiffness to obtain a lumped parameter model of the building structure. Based on the lumped parameter model of the building structure, a graph network structure is constructed. Each structural node is regarded as a vertex in the graph, and the component connection relationship is regarded as an edge in the graph to obtain a structural topology graph network. The structural topology network is processed with edge weight assignment. By calculating the stiffness transfer relationship between adjacent nodes, the bending stiffness coefficient and shear stiffness ratio are mapped to edge weight values ​​to obtain a weighted structural graph network. Based on the weighted structural graph network, message passing calculation is performed to update the structural feature attributes of each node through information exchange between nodes to obtain structural node feature data; Perform vibration mode calculation on the characteristic data of the structural nodes, obtain the vibration period and vibration mode of the structure by solving the characteristic value equation, and obtain the dynamic characteristic data of the simplified model; The component layout optimization is performed based on the dynamic characteristic data of the simplified model. The dynamic characteristics of the simplified model are matched with the target characteristics by adjusting the node positions and component connection methods to obtain the structural layout prediction results.

[0033] Specifically, the structural stiffness distribution scheme is simplified by transforming the complex three-dimensional structural system into a Timoshenko beam model consisting of multiple distributed mass and distributed stiffness elements, while ensuring the fidelity of the mechanical properties. This model is distinguished by simultaneously considering the structural response under bending and shear deformation, unlike traditional simplified models that only consider bending. During the construction process, the spatial geometry, material properties, connection nodes, and spatial arrangement of each beam and column are extracted from the building information model. Each continuous beam or column segment is then divided into discrete elements according to a predefined equivalence rule. Each element is assigned a stiffness value and a mass value, corresponding to two dominant parameters in the structural load characteristics. The entire building structure is then transformed into a lumped parameter model consisting of discrete component elements, each with a clear physical definition and suitable for subsequent graphical model construction. Based on this lumped model, the next key step is to construct a graph network structure. Each node in the graph corresponds to a component connection point, namely, the connection between beams and columns, columns and foundations, or beams in the model. An edge is created between each pair of nodes with a direct physical connection, indicating the existence of a force transmission path between the components. This step requires traversing the entire lumped model, identifying the connection points of each component in space, and recording the node numbers of its starting and ending points. The resulting topology is a directed graph, where the connections between nodes fully reflect the connections between actual components. There are no redundant edges, and the direction of the edges corresponds to the component installation direction, ensuring data consistency during subsequent dissemination.

[0034] Each edge in the graph is then assigned a weight. This operation aims to quantitatively map stiffness information from structural mechanics into the graph structure, making it physically interpretable. The weight of each edge is determined by the bending and shear stiffness of the components connected by adjacent nodes. Bending stiffness reflects a component's ability to resist deformation, while shear stiffness describes its resistance to deformation under horizontal forces. When calculating edge weights, the cross-sectional dimensions, material modulus, and geometric parameters of the components connected by the two end nodes are extracted, and their stiffness performance under load is calculated. This result is used as the edge weight. The resulting weighted graph not only possesses a topological structure but also incorporates quantitative parameters reflecting physical properties, laying the foundation for subsequent information dissemination. Executing message-passing computations based on a weighted graph network structure is key to integrating graph neural networks into structural physics modeling. During this stage, each node is initialized with multidimensional attributes such as the corresponding component type, geometric dimensions, material type, load type, and spatial position. These attributes are encoded as vectors. The graph neural network transmits feature information between adjacent nodes according to predefined propagation rules. Edge weights determine the degree and direction of information transfer. After each round of propagation, a node updates its eigenvector, reflecting its position within the overall structure and the changing trends in its mechanical behavior. After multiple rounds of information propagation, each node's eigenvector ultimately contains not only its own information but also incorporates changes in the characteristics transmitted by neighboring nodes, reflecting the mechanical interactions within the overall structural arrangement.

[0035] After message propagation is complete, mode shape calculations are performed on the node characteristic data to extract the overall dynamic characteristics of the structure. This stage primarily extracts the changing trends of each node's eigenvalues ​​to identify the node's natural vibration period and corresponding mode shape. The node's vibration period is determined by its stiffness and mass, while the mode shape describes the distribution of the structure's overall vibration patterns at different frequencies. This process, based on the node characteristics in the graph, derives the global dynamic response pattern through frequency domain analysis of the entire graph.

[0036] Based on this mode shape information, component layout optimization is performed. The optimization goal is to adjust the structural layout to make its vibration characteristics more reasonable and avoid problems such as local mode shape concentration and overall mode shape mutation. This step adjusts the position of the nodes, that is, changes the layout of the components in space, and optimizes the connection method of the edges, that is, modifies the connection order or connection method of the components, so that the energy distribution of the overall structure under each main mode is more uniform and the vibration response is more stable. The optimization algorithm uses a combination of greedy search and constraint condition verification. The mode shape characteristics of the graph are recalculated after each adjustment. When the set target characteristics are met, the output is the structural layout prediction result.

[0037] In a specific embodiment, the process of executing step S104 may specifically include the following steps: The structural layout prediction results are classified into component types. When a component is mainly subjected to axial force, it is determined to be a column component; when a component is mainly subjected to bending moment, it is determined to be a beam component; when a component is subjected to large axial force and bending moment at the same time, it is determined to be a beam-column combination component, and component classification data is obtained; Based on the component classification data, standard component library matching is performed. If the component's stress characteristics meet the national standard specifications, the standard section is selected. If the component's stress characteristics exceed the standard range, a customized section is generated to obtain component parametric description data. Perform geometric modeling on the component parametric description data. If the component is a beam, an I-section or box section is generated. If the component is a column, an H-section or circular tube section is generated. If the component is a wall, a plate section is generated to obtain a detailed component model. Node connection processing is performed based on the detailed component model. When beams and columns intersect, a rigid connection is formed. When supports intersect the main structure, a hinged connection is formed. When different structural units are connected, a transition connection is formed. Node design data is obtained. Assemble the node design data and component detailed models as a whole to obtain a preliminary building structure model; Based on the preliminary building structure model, the structural details are improved. Stiffening ribs are added when the nodes are under concentrated force. Connection reinforcement plates are set when the stress at the connection part is large. Anchor devices are designed at the connection between the foundation and the superstructure to obtain the building structure model.

[0038] Specifically, the structural layout prediction results are used as input, which include the three-dimensional coordinates of each node in space, the starting and ending nodes of the component, the force vector and moment data. Logical judgment is performed on these data to extract the axial force component and bending moment component of each component on the force coordinate axis, normalize their sizes, and then calculate the ratio of axial force to bending moment. If the ratio is significantly biased towards axial force, it is calibrated as a column component; if the bending moment is dominant, it is calibrated as a beam component; if both are the main influencing factors, it is classified as a beam-column composite component. The classification process adopts a rule mapping method. All components are automatically assigned component type labels according to this logic and written into the component classification table. The table fields include component number, starting and end node numbers, main force type, geometric information source, etc., providing a classification basis for subsequent modeling steps.

[0039] The component library matching phase then begins. Based on the component classification table, key parameters such as force properties, dimensions, span length, and load level are compared with the national standard component library. The standard component library is a pre-organized database of national design standards, containing various specifications of H-section steel, box steel, circular tube sections, C-section steel, and other types, along with their applicable mechanical ranges and boundary conditions. The matching logic is as follows: if the span length and load level of the current component completely fall within the data range of a standard component, that section type is directly selected as the matching result. If the span length is too large or the bending moment exceeds the standard range, the component is marked as requiring a custom section and enters the section generation module. Based on the maximum internal force and the component slenderness ratio, a section size that meets the strength and stiffness requirements is generated. All matching results are written into a component parametric description data table, which includes fields such as component type, section form, dimensional parameters, material grade, and connection constraints. Once the component parameter data is determined, the geometric modeling phase begins, where a rule-driven modeling program generates a 3D model based on the parameter data. For beams, I-sections or box sections are generated based on the load path and usage. I-sections are used for medium-span areas subject primarily to bending, while box sections are used for areas with concentrated forces requiring increased torsional stiffness. For columns, H-sections are generated to withstand combined bidirectional bending moments, while circular tube sections are used for areas requiring high torsional resistance and where a uniform force field is required for node connections. For walls, planar plates with thickness parameters are generated, with section height and thickness derived from a parameter table. The generated model includes complete 3D geometry, local normals, boundary control points, and pre-set connection point information, ensuring direct connection to node components.

[0040] After the component model is generated, its connection method needs to be processed by node design. Matching is performed according to the connection method rules recorded in the component classification table. For example, if the intersection of the beam and column components is a rigid connection, a rigid constraint feature is generated at the connection position of the component model; if the diagonal brace is connected to the main beam, a hinged connection is formed, and a single-degree-of-freedom rotation mechanism is set at the corresponding node; if the connection between different structural units is completed, such as the intersection of the platform structure and the main factory frame, it is defined as a transition connection. The interface section is designed according to the cross-sectional form of the two components, and connecting transition components such as transition plates or flexible connectors are inserted at the same time. The node design data structure includes connection type, node number, connection component number, constraint type, structural reinforcement requirements, etc., forming a complete connection information table.

[0041] After the node data and component models are complete, the overall assembly process begins, assembling all components according to their spatial positions and connection relationships into a preliminary model of the overall building structure. Using the node table as a control index, each component model is inserted into the main model structure by number. Boolean operations or parametric connection modules are used to establish the connection between components, forming a complete three-dimensional spatial skeleton model.

[0042] After obtaining the preliminary structural model, the structural details are refined, mainly focusing on local modeling and reinforcement of the nodes and component details in the areas where stress is concentrated. The judgment criteria are as follows: if a node is subject to concentrated stress and has a large number of connected components, stiffeners are added to the node to ensure that the node connection does not yield locally; if the connection is a force path conversion point, such as the intersection of beams and columns, or the connection between trusses and roof beams, reinforcement plates are installed to disperse stress; if there is concentrated force transmission at the intersection of the foundation component and the superstructure, anchoring devices such as embedded parts, anchor bolts, or anchor bars are set in the connection area to achieve effective transition of component forces. The above-mentioned reinforcement components are also modeled in a parametric form, and the component information table and node connection table are added to maintain data consistency.

[0043] In a specific embodiment, the process of executing step S105 may specifically include the following steps: Define load conditions for the building structure model. Create different design conditions by combining dead loads, live loads, wind loads, and seismic loads. Consider the P-Delta effect when both vertical and horizontal loads are present, and obtain a multi-condition load combination scheme. Static analysis is performed based on a multi-load combination scheme. Elastic analysis is performed when the structure is in normal use, and nonlinear analysis is performed when the structure is close to the limit state to obtain the structural internal force distribution data. Perform component cross-section design on the structural internal force distribution data. When the component is subjected to axial force, the cross-sectional area is optimized; when the component is subjected to bending moment, the cross-sectional inertia moment is optimized; when the component is subjected to shear force, the web thickness is optimized to obtain the optimized component cross-sectional data. Node connection verification is performed based on optimized component section data. When the node load is simple, standard connection is used. When the node load is complex, detailed finite element analysis is performed to obtain node connection design data. Displacement verification is performed on the node connection design data and the optimized component cross-section data. When the horizontal displacement exceeds the limit, the lateral stiffness is increased. When the vertical deflection exceeds the limit, the component cross-section is adjusted. When the local deformation of the component is too large, stiffness reinforcement is added to obtain a structural deformation control plan. The overall performance of the structure is evaluated based on the structural deformation control scheme. When the structural construction is difficult, the node design is simplified to obtain the structural design scheme.

[0044] Specifically, load cases are defined for the building structure model, with data input sourced from the building geometry and generated component parameter sets. The system automatically constructs data for dead loads, live loads, wind loads, and seismic loads based on model component material, dimensions, spatial location, and usage. Dead loads are calculated from component volume and material density, live loads are mapped from typical values ​​in a library of building function standards, wind loads are segmented based on building height and wind pressure zone standards, and seismic loads are derived from a response spectrum library by matching the building period with the design intensity level. Each load is applied to component nodes as a three-dimensional vector, and multiple combinations are generated based on the load case combination logic, such as dead load plus live load, dead load plus wind load plus seismic load, and so on. For combinations involving both vertical and horizontal loads, the P-Delta effect iterative analysis mode is also enabled, which considers the additional force effects caused by structural displacement in the calculation to ensure the integrity of the structural stability analysis. The data output at this stage is the structural input load matrix under multiple load combination scenarios, providing a basis for subsequent structural response analysis.

[0045] Next, static analysis is carried out, and different analysis models are used according to different working condition combinations. When the working condition is in the normal use stage of the building, elastic analysis is carried out to extract the linear response of the component under load, including axial force, bending moment, shear force and torque, etc.; when the working condition is close to the limit state, such as considering the action of a large earthquake or the ultimate load of strong wind, the nonlinear analysis stage is entered to consider nonlinear responses such as material yielding, large deformation of components and connection failure. The elastic analysis uses the matrix displacement method or the finite element method to discretize the structure into nodes and units, and solves the deformation response of each node and the internal force response of each unit; the nonlinear analysis introduces factors such as component stiffness weakening, plastic hinge formation and node rotation, and applies iterative corrections to the component units until the convergence conditions are met. The analysis results are output as a complete internal force distribution data set for each component under each working condition, including the maximum internal force position, stress value and deformation. All data are indexed into a table with the structural unit number and the working condition number to support subsequent optimization.

[0046] When performing cross-sectional design processing on the structural internal force distribution data, the maximum internal force data of each component is first read and grouped according to the dominant force type. For components dominated by axial forces, the minimum cross-sectional area is calculated based on compressive stress and stability; for components dominated by bending, the required moment of inertia is calculated; for components dominated by shear, the web thickness is optimized based on the shear yield of the web. This process relies on the national steel structure design standard database. By looking up tables or using analytical functions to compare the safety factors of different sections under specified internal forces, the optimal section is selected and written into the component attribute table. If the current section meets the requirements and is a standard component, the original design is maintained; if not, the section is upgraded or the section form is adjusted, thereby forming a component section optimization data set.

[0047] Based on the optimized data of the component cross-section, the node connection verification process is further carried out, and the connection scheme is matched according to the internal force value of the connection and the component type. For nodes with uniform internal force distribution and simple force, standard connection templates are directly used, such as welded rigid connection, bolted hinged connection and other standard schemes; while for nodes with complex internal forces, overlapping bending and shear, or intersection of multiple components, it is necessary to import a detailed finite element analysis module, establish a local model of the node, input the node component load, boundary conditions and material properties, conduct local stress and strain analysis of the node, identify weak areas, adjust the connection plate size and weld length, and backfill the node connection design data into the main structure model. Then enter the displacement verification process stage. With the goal of structural deformation control, read the node displacement results of the overall structural model, and compare the displacement limit standard to determine whether it exceeds the limit. If the structure's horizontal drift under wind loads exceeds the code allowable value, the structural lateral stiffness needs to be increased, such as by adding bracing members or strengthening the frame. If the mid-span deflection of a beam exceeds the limit, the beam cross-section is optimized to increase stiffness. If excessive local deformation is observed at certain nodes, local reinforcement is implemented, such as by adding stiffeners, doubling components, or varying cross-sections. The output of this stage is a structural deformation control plan table, which records the numbers of all components requiring optimization or reinforcement, along with the optimization measures and recommendations. Based on all optimization results, an overall structural performance assessment is conducted. This process comprehensively scores the number of structural components, node complexity, and accessibility, and evaluates the feasibility of the structural solution based on structural stability, force rationality, and construction economy. If the construction difficulty score is high, the system automatically determines whether the node structure is too complex. For example, if there are multiple component special-shaped connections or too many anchoring devices, recommendations are made to simplify the node design, adjust the component connection logic, or standardize cross-section standards to achieve a balance between structural performance and construction efficiency. The final output is a structural design plan data file, including all component numbers, cross-section types, connection methods, and control parameters.

[0048] In a specific embodiment, the process of executing step S106 may specifically include the following steps: Extract the design description of the structural design scheme and obtain the structural design description document by arranging the design basis, load data and calculation results; Based on the structural design plan, construction drawings are drawn and processed to obtain a set of structural construction drawings by generating plan views, cross-section views and node details; Conduct material statistics processing on the structural design scheme and obtain a material list by calculating the quantity and specifications of various components; Arrange the construction sequence based on the structural construction drawing set, and obtain the construction guidance document by determining the component installation sequence and construction technical requirements; Perform data association processing on the relevant files of the structural design scheme, and obtain a database of the corresponding relationship between design parameters and analysis results by recording the mapping relationship between design parameters and calculation results; Based on the corresponding database of design parameters and analysis results, integration processing is carried out, and the building structure design scheme is obtained by integrating various design documents and construction information.

[0049] Specifically, during the design specification extraction phase, the automatic document generation module is invoked based on the component numbers, cross-section types, connection methods, calculation control conditions, and node force data in the structural design solution database. The module then automatically populates the design specification entries using pre-set templates. During the data processing process, basic design basis items (such as load specifications, steel structure design specifications, and seismic fortification standards) are read from the template library and inserted into the document header. Subsequently, a table of component mechanical data and load case combinations is accessed to provide a typical description for each component type, including the maximum load case number, control internal force value, corresponding calculation section, and construction type. For example, for a bent main beam, the description clearly states that the design is based on a combination of "constant + live + wind" loads, the maximum bending moment occurs at mid-span, the selected section is a box steel section, the material grade is Q355B, and the connection type is rigid. All content is automatically filtered and matched based on the previous analysis results and parameter tables, eliminating manual entry and ensuring accuracy and consistency. Once the design specification document is output, the construction drawing process immediately begins. The input data of this stage is the spatial geometric model of the component, the connection node table, the construction detail model and the section template library. The output content includes construction drawings such as plan views, elevation views, section views and node details. The drawing generation process first projects the main frame model of the structure, identifies the main direction of each component, and generates a plan view in a two-dimensional coordinate system; the elevation view is achieved by projecting the component on the XZ or YZ plane. When drawing the section view, the component cross-sectional geometric parameters are extracted from the model data, and the standard I-type, box-type, H-type or circular tube cross-sectional profile is drawn according to the component classification; the node details are extracted from the node connection table. The connection parts, bolt positions, weld symbols and reinforcement component data are zoomed in to generate a local structural drawing. After all the graphics are generated, they are uniformly numbered and written into the drawing index table to achieve a one-to-one correspondence between the drawings and the models.

[0050] After the drawings are completed, the structural design plan undergoes material statistics. Based on the component parametric description table and node connection data, the cross-sectional specifications, lengths, quantities, and total weight of each component type are calculated. Material breakdowns are then output in batches. During this process, all components are grouped by type, and the number of components of the same specification is aggregated and multiplied by the material volume and density of each component to determine the total material volume. For connection materials such as bolts and welding rods, the number, type, and length of connectors are extracted from the node connection table and then compared against a standard library. The output is a material breakdown, including steel classification (H-beam, box steel, circular pipe, etc.), connection material type (high-strength bolts, welding rods), and auxiliary materials (anchor plates, stiffeners), along with their specifications, quantities, and standard codes. This is used for procurement planning and cost control. A construction sequence document is generated based on the drawings and material data, serving as the basis for subsequent construction organization. The construction sequence is organized based on component number, spatial location, construction node coordinates, and installation height data, and is arranged according to the installation path from the starting point to the end point. The need for temporary supports is also determined based on the structural self-stability. The process uses a topological sorting algorithm to analyze component connection diagrams, identify component installation dependencies, assign component partition numbers, and output component installation batch lists and lifting numbers. Construction technical requirements are automatically generated based on the component installation method and connection node type. For example, when a component is a double-jointed box beam with welded end nodes, the technical requirements automatically include items such as welding sequence, controlling weld deformation, and node retesting. The final output is a construction instruction document, which includes component numbers, installation sequence, connection method, construction process, and precautions.

[0051] Data linkage across various structural design documents is crucial for ensuring data consistency. This process maps structural parameters (such as cross-sectional dimensions, component numbers, and connection methods), calculation results (internal forces, displacements, and load conditions), design drawings, and bills of materials using unified numbers. Using component numbers as the primary key, a relational database is constructed, recording component-specific design parameters, construction drawing locations, node detail numbers, and structural analysis results. This allows any drawing or component to be traced back to the original design basis and calculation sources. This database is structured and stored in a model management platform with an interface to support subsequent construction management and project archiving. By integrating all design parameter data, construction drawings, material lists, and construction specifications, a unified building structural design output package is created. This package includes design specification documents, construction drawing sets, material statistics, construction instructions, and a parameter analysis database. All content is bidirectionally linked to the data tables using unique component numbers, ensuring a closed-loop information loop with clear structure, traceability, reviewability, and actionability.

[0052] The above describes the method for automated design of building structures based on artificial intelligence in the embodiment of the present application. The following describes the automated design system for building structures based on artificial intelligence in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the building structure automatic design system based on artificial intelligence includes: The acquisition module 201 is used to collect data on the architectural design task and obtain a building information model containing collaborative information; A calculation module 202 is used to perform load analysis and stiffness characteristic parameter calculation on the building information model, and obtain a structural stiffness distribution scheme by determining the bending stiffness coefficient and the shear stiffness ratio; A construction module 203 is used to construct a simplified model of a lumped Timoshenko beam based on the structural stiffness distribution scheme and integrate a graph network optimization algorithm to obtain a structural layout prediction result through component relationship analysis and mode shape matching calculation; An input module 204 is used to input the structural layout prediction results into a parametric modeling program to generate components and obtain a building structure model; An analysis module 205 is used to perform multi-condition analysis and calculation on the building structure model to obtain a structural design solution; The generation module 206 is used to generate design documents and construction information based on the structural design scheme, and establish a database of corresponding relationships between design parameters and analysis results to obtain a building structure design scheme.

[0053] Through the collaborative collaboration of these various components, data collection for architectural design tasks is performed, resulting in a building information model containing collaborative information. This allows information such as building geometry, equipment layout, material parameters, and load application locations to be stored and accessed in a unified data format, effectively resolving the issues of multi-disciplinary data disconnection and repeated modeling that exist in traditional design processes. By performing load analysis and calculating stiffness characteristic parameters on this building information model, and further determining the bending stiffness coefficient and shear stiffness ratio, a stiffness response distribution dataset for the entire structural domain is constructed, improving the overall force balance of the structure and the ability to identify local responses. On this basis, a simplified model of a lumped Timoshenko beam is constructed using the structural stiffness distribution scheme, and a graph network optimization algorithm is integrated to map the original spatial component relationships into a graph network with physical properties. Through component relationship analysis and mode matching calculations, the prediction, deduction, and adjustment optimization of the structural layout are achieved, significantly improving the rationality of the structural layout and its seismic performance. The predicted structural layout results are then input into a parametric modeling program for component generation. This intelligently determines component classification based on the type of force, automatically selects standard or customized sections, and quickly generates calculable and plottable three-dimensional component models, significantly reducing modeling workload and improving consistency. Next, by performing multi-condition analysis and calculations on the generated building structure model, a comprehensive assessment of structural stiffness, displacement, and component internal forces is completed under multiple load combination scenarios. Component size and connection methods are adjusted based on the response results to ensure structural safety and economy. Finally, the structural design solution automatically generates design documents and construction information, establishes a database of corresponding relationships between design parameters and analysis results, completes structured archiving and data integration of design results, and implements a closed-loop process from inputting architectural design tasks to outputting complete structural design solutions.

[0054] above Figure 2 The artificial intelligence-based automated building structure design system in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The artificial intelligence-based automated building structure design device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0055] Figure 3The figure is a schematic diagram of the structure of an AI-based automated building structure design device provided by an embodiment of the present invention. The AI-based automated building structure design device 300 may vary significantly depending on configuration or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors), memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage media 330 may be either transient or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions and operations within the AI-based automated building structure design device 300. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, allowing the AI-based automated building structure design device 300 to execute the series of instructions and operations stored in the storage medium 330 to implement the steps of the aforementioned AI-based automated building structure design method.

[0056] The artificial intelligence-based building structure automation design device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The structure of the artificial intelligence-based building structure automation design equipment shown does not constitute a limitation on the artificial intelligence-based building structure automation design equipment provided by the present invention, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0057] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the method for automated design of building structures based on artificial intelligence.

[0058] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0059] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling an artificial intelligence-based automated building structure design device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0060] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An artificial intelligence-based automated building structure design method, characterized in that: The method comprises: Collect data for architectural design tasks and obtain a building information model containing collaborative information; Performing load analysis and stiffness characteristic parameter calculation on the building information model, and obtaining a structural stiffness distribution scheme by determining a bending stiffness coefficient and a shear stiffness ratio; Based on the structural stiffness distribution scheme, a simplified lumped Timoshenko beam model was constructed and integrated with a graph network optimization algorithm. The structural layout prediction results were obtained through component relationship analysis and mode shape matching calculation. Inputting the structural layout prediction results into a parametric modeling program to generate components and obtain a building structure model; Performing multi-condition analysis and calculation on the building structure model to obtain a structural design scheme; Based on the structural design scheme, design documents and construction information are generated, and a database of corresponding relationships between design parameters and analysis results is established to obtain a building structure design scheme.

2. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: The data collection for the architectural design task is performed to obtain a building information model containing collaborative information, including: Conduct project demand analysis on architectural design tasks to obtain basic architectural design parameters; Conducting on-site condition surveys on the basic parameters of the building design, and obtaining building space point cloud data by measuring the site topography and surrounding environmental constraints; Performing feature extraction and contour recognition processing on the building space point cloud data, and obtaining a building space geometric model by separating the spatial information of the building main structure and non-structural components; Perform professional data association processing on the architectural space geometric model, import the layout drawings and technical parameter tables of various professional equipment, and obtain multi-professional collaborative data; Based on the multi-disciplinary collaborative data, load identification and classification are performed, and the structural load distribution plan is obtained by calculating the weight of each device and determining its position in the structure; The structural load distribution scheme is integrated to obtain a building information model containing collaborative information.

3. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: The load analysis and stiffness characteristic parameter calculation of the building information model are performed, and a structural stiffness distribution scheme is obtained by determining a bending stiffness coefficient and a shear stiffness ratio, including: Performing structural type identification processing on the building information model, and obtaining building structural system data by analyzing the building plan layout and vertical component distribution characteristics; Performing load sub-item calculation based on the building structure system data, by separating dead load, live load, wind load and earthquake load and determining their action directions, thereby obtaining multi-type load action data; Performing load combination processing on the multi-type load action data, and obtaining critical load condition data by combining load effects under different working conditions according to design specification requirements; Calculating and processing the structural bending stiffness based on the critical load condition data, and obtaining the bending stiffness coefficient by analyzing the resistance of the structure under bending deformation; Performing shear deformation analysis on the building information model, and obtaining a shear stiffness ratio by calculating shear deformation characteristics of the structure under horizontal force; The stiffness distribution optimization process is performed based on the bending stiffness coefficient and the shear stiffness ratio, and a structural stiffness distribution scheme is obtained by balancing the stiffness distribution of each region and considering special treatment of the stiffness mutation region.

4. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: Based on the structural stiffness distribution scheme, a simplified lumped Timoshenko beam model is constructed and integrated with a graph network optimization algorithm. Through component relationship analysis and mode shape matching calculation, the structural layout prediction results are obtained, including: The structural stiffness distribution scheme is simplified, and a lumped parameter model of the building structure is obtained by equating the complex building structure to a Timoshenko beam element with distributed mass and stiffness. Constructing a graph network structure based on the building structure lumped parameter model, obtaining a structural topology graph network by treating each structural node as a vertex in the graph and the component connection relationship as an edge in the graph; Performing edge weight assignment processing on the structural topology graph network, mapping the bending stiffness coefficient and the shear stiffness ratio into edge weight values ​​by calculating the stiffness transfer relationship between adjacent nodes, and obtaining a weighted structural graph network; Perform message passing calculation based on the weighted structure graph network, update the structural feature attributes of each node through information exchange between nodes, and obtain structural node feature data; Performing vibration mode calculation on the structural node characteristic data, obtaining the vibration period and vibration mode of the structure by solving the characteristic value equation, and obtaining simplified model dynamic characteristic data; Component layout optimization processing is performed based on the simplified model dynamic characteristic data, and the dynamic characteristics of the simplified model are matched with the target characteristics by adjusting the node positions and component connection methods to obtain a structural layout prediction result.

5. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: The step of inputting the structural layout prediction results into a parametric modeling program to generate components and obtain a building structure model comprises: The structural arrangement prediction result is subjected to component type classification processing, where a component is determined to be a column component when it is mainly subjected to axial force, a beam component when it is mainly subjected to bending moment, and a beam-column combination component when it is simultaneously subjected to large axial force and bending moment, thereby obtaining component classification data; Performing standard component library matching processing based on the component classification data, selecting a standard section if the component's stress characteristics meet national standard specifications, and generating a customized section if the component's stress characteristics exceed the standard range to obtain component parametric description data; Performing geometric modeling processing on the parametric description data of the component, generating an I-section or a box section when the component is a beam type, generating an H-section or a circular tube section when the component is a column type, and generating a plate section when the component is a wall, to obtain a detailed model of the component; Node connection processing is performed based on the detailed model of the component, forming a rigid connection when beams and columns intersect, forming a hinged connection when supports intersect the main structure, and forming a transition connection when different structural units connect, to obtain node design data; Assembling the node design data and the component detailed model as a whole to obtain a preliminary building structure model; Based on the preliminary building structure model, the structural details are improved. When the nodes are subjected to concentrated force, stiffening ribs are added. When the stress at the connection part is large, connection reinforcement plates are set. When the foundation and the superstructure are connected, anchoring devices are designed to obtain the building structure model.

6. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: The multi-condition analysis and calculation of the building structure model to obtain a structural design scheme includes: Defining load conditions for the building structure model, forming different design conditions by combining dead loads, live loads, wind loads, and seismic loads, and considering the P-Delta effect when both vertical and horizontal loads exist, to obtain a multi-condition load combination scheme; Perform static analysis based on the multi-load combination scheme, perform elastic analysis when the structure is in normal use, and perform nonlinear analysis when the structure is close to the limit state to obtain structural internal force distribution data; Performing component cross-section design processing on the structural internal force distribution data, optimizing the cross-sectional area when the component is subjected to axial force, optimizing the cross-sectional moment of inertia when the component is subjected to bending moment, and optimizing the web thickness when the component is subjected to shear force, to obtain component optimized cross-sectional data; Perform node connection verification based on the optimized cross-section data of the component. When the node is subjected to simple forces, a standard connection is used. When the node is subjected to complex forces, a detailed finite element analysis is performed to obtain node connection design data. Performing displacement verification processing on the node connection design data and the optimized cross-section data of the component, increasing the lateral stiffness when the horizontal displacement exceeds the limit, adjusting the component cross-section when the vertical deflection exceeds the limit, and increasing stiffness reinforcement when the local deformation of the component is too large, thereby obtaining a structural deformation control plan; An overall structural performance evaluation is performed based on the structural deformation control scheme. When the structural construction is difficult, the node design is simplified to obtain a structural design scheme.

7. The method for automated building structure design based on artificial intelligence according to claim 1, characterized in that: Generating design documents and construction information based on the structural design scheme, and establishing a database of corresponding relationships between design parameters and analysis results to obtain a building structure design scheme includes: Extracting the design description of the structural design scheme, and obtaining a structural design description document by arranging the design basis, load data and calculation results; Performing construction drawing processing based on the structural design scheme, and obtaining a set of structural construction drawings by generating plan views, cross-section views, and node details; Perform material statistics on the structural design scheme and obtain a material list by calculating the quantity and specifications of various components; Arrange the construction sequence based on the structural construction drawing set, and obtain a construction guidance document by determining the component installation sequence and construction technical requirements; Performing data association processing on relevant files of the structural design scheme, and obtaining a database of corresponding relationships between design parameters and analysis results by recording the mapping relationship between design parameters and calculation results; Based on the corresponding relationship database between the design parameters and the analysis results, integration processing is performed, and the building structure design scheme is obtained by integrating various design documents and construction information.

8. An artificial intelligence-based automated building structure design system, characterized in that: For implementing the method for automated building structure design based on artificial intelligence according to any one of claims 1 to 7, the automated building structure design system based on artificial intelligence comprises: The acquisition module is used to collect data for architectural design tasks and obtain a building information model containing collaborative information; a calculation module, configured to perform load analysis and stiffness characteristic parameter calculation on the building information model, and obtain a structural stiffness distribution scheme by determining a bending stiffness coefficient and a shear stiffness ratio; A construction module is used to construct a simplified model of a lumped Timoshenko beam based on the structural stiffness distribution scheme and integrate a graph network optimization algorithm to obtain a structural layout prediction result through component relationship analysis and mode shape matching calculation; An input module, configured to input the structural arrangement prediction results into a parametric modeling program to generate components and obtain a building structure model; An analysis module is used to perform multi-condition analysis and calculation on the building structure model to obtain a structural design solution; The generation module is used to generate design documents and construction information based on the structural design scheme, and establish a database of corresponding relationships between design parameters and analysis results to obtain a building structure design scheme.

9. An artificial intelligence-based automated building structure design device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the method for automated design of building structures based on artificial intelligence according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the processor is enabled to execute the artificial intelligence-based automated building structure design method according to any one of claims 1 to 7.

Images