A modular steel structure connection method and system based on BIM model

Abstract

A modular steel structure connection method and system based on a BIM model. It creates different connection schemes for modular steel structures in the BIM model, and introduces artificial intelligence-based data processing and analysis algorithms at the back end to analyze the stress distribution of different connection schemes, thereby achieving more accurate connection scheme recommendations. In this way, it is possible to perform intelligent evaluation of the stress distribution in different modular steel structure connection schemes to automatically recommend steel structure connection schemes, which not only improves the application intelligence of BIM, but also provides decision support for the connection of modular steel structures.

Description

Technical Field

[0001] The present disclosure relates to the field of modular buildings, and in particular to a modular steel structure connection method and system based on a BIM model. Background Art

[0002] Modular construction is an increasingly popular construction method that improves the efficiency, quality and sustainability of construction projects by breaking down buildings into individual modules, producing them in factories and finally assembling them on site.

[0003] As a commonly used form of building structure, steel structure has the advantages of high strength, fast construction speed and light weight. In modular buildings, the design and selection of steel structure connections are crucial to the safety and stability of the entire building. However, the connection design of modular steel structures is a complex process. Engineers need to consider a variety of factors, including connection type and stress distribution.

[0004] Therefore, an optimized modular steel structure connection solution is desired. Summary of the invention

[0005] The present disclosure is made in view of the above problems. One object of the present disclosure is to provide a modular steel structure connection method and system based on a BIM model.

[0006] An embodiment of the present disclosure provides a modular steel structure connection method based on a BIM model, which includes:

[0007] Create a first connection scheme of modular steel structure and a second connection scheme of modular steel structure in the BIM model;

[0008] Using finite element analysis to obtain stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain a first stress distribution diagram and a second stress distribution diagram;

[0009] Performing local stress mode characteristic analysis on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors;

[0010] Constructing a stress distribution adjacency characteristic matrix of the sequence of the first local stress distribution characteristic vectors and constructing a stress distribution adjacency characteristic matrix of the sequence of the second local stress distribution characteristic vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix;

[0011] Performing feature extraction on the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively by a stress distribution topology feature extractor based on a deep neural network model to obtain a first stress distribution adjacency topology feature matrix and a second stress distribution adjacency topology feature matrix;

[0012] Performing a graph-based local stress distribution feature interactive analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix to obtain a first-solution global stress distribution feature matrix, and performing a graph-based local stress distribution feature interactive analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix to obtain a second-solution global stress distribution feature matrix; and

[0013] A feature comparison is performed on the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0014] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the first stress distribution diagram and the second stress distribution diagram are subjected to local stress mode characteristic analysis to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors, including:

[0015] Performing image block processing on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution maps and a sequence of second local stress distribution maps; and

[0016] The sequence of the first local stress distribution maps and the sequence of the second local stress distribution maps are respectively passed through a stress local pattern feature extractor based on a convolutional neural network model to obtain a sequence of the first local stress distribution feature vectors and a sequence of the second local stress distribution feature vectors.

[0017] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the deep neural network model is a convolutional neural network model.

[0018] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a first scheme global stress distribution feature matrix, and the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a second scheme global stress distribution feature matrix, including:

[0019] The sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix, as well as the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix are respectively passed through a local stress distribution feature interactor based on a graph convolutional neural network model to obtain the first scheme global stress distribution feature matrix and the second scheme global stress distribution feature matrix.

[0020] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, a feature comparison is performed on the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including:

[0021] Calculating a differential characteristic matrix between the global stress distribution characteristic matrix of the first solution and the global stress distribution characteristic matrix of the second solution; and

[0022] The differential feature matrix is ​​passed through a classifier-based solution evaluator to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection solution of the modular steel structure is better than the second connection solution of the modular steel structure.

[0023] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the differential feature matrix is ​​passed through a scheme evaluator based on a classifier to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including:

[0024] Expanding the differential feature matrix into classification feature vectors according to row vectors or column vectors;

[0025] Performing fully connected encoding on the classification feature vector using the fully connected layer of the classifier-based solution evaluator to obtain an encoded classification feature vector; and

[0026] The encoded classification feature vector is input into the Softmax classification function of the classifier-based solution evaluator to obtain the evaluation result.

[0027] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, it also includes a training step: for training the stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topological feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based scheme evaluator.

[0028] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the training step includes:

[0029] Acquire training data, wherein the training data includes a first training connection scheme of a modular steel structure and a second training connection scheme of a modular steel structure created in a BIM model;

[0030] Using finite element analysis to obtain stress distribution diagrams of the first training connection scheme of the modular steel structure and the second training connection scheme of the modular steel structure to obtain a first training stress distribution diagram and a second training stress distribution diagram;

[0031] Performing image block processing on the first training stress distribution map and the second training stress distribution map to obtain a sequence of first training local stress distribution maps and a sequence of second training local stress distribution maps;

[0032] Passing the sequence of the first training local stress distribution maps and the sequence of the second training local stress distribution maps through the stress local pattern feature extractor based on the convolutional neural network model to obtain a sequence of the first training local stress distribution feature vectors and a sequence of the second training local stress distribution feature vectors respectively;

[0033] Constructing a stress distribution adjacency feature matrix of the first training sequence of local stress distribution feature vectors and constructing a stress distribution adjacency feature matrix of the second training sequence of local stress distribution feature vectors to obtain a first training stress distribution adjacency matrix and a second training stress distribution adjacency matrix;

[0034] The stress distribution topology feature extractor based on the convolutional neural network model performs feature extraction on the first training stress distribution adjacency matrix and the second training stress distribution adjacency matrix respectively to obtain a first training stress distribution adjacency topology feature matrix and a second training stress distribution adjacency topology feature matrix;

[0035] The sequence of the first training local stress distribution feature vectors and the first training stress distribution adjacency topology feature matrix, as well as the sequence of the second training local stress distribution feature vectors and the second training stress distribution adjacency topology feature matrix are respectively passed through the local stress distribution feature interactor based on the graph convolutional neural network model to obtain the first training scheme global stress distribution feature matrix and the second training scheme global stress distribution feature matrix;

[0036] Calculating a training difference feature matrix between the global stress distribution feature matrix of the first training scheme and the global stress distribution feature matrix of the second training scheme;

[0037] Optimizing the training differential feature matrix to obtain a training optimized differential feature matrix;

[0038] Passing the training optimized differential feature matrix through the classifier-based solution evaluator to obtain a classification loss function value; and

[0039] The stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topology feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based solution evaluator are trained based on the classification loss function value.

[0040] For example, according to the modular steel structure connection method based on the BIM model of an embodiment of the present disclosure, the training differential feature matrix is ​​optimized to obtain a training optimized differential feature matrix, including:

[0041] Clustering each eigenvalue of the training differential feature matrix based on the distance between eigenvalues ​​to obtain a cluster feature set; and

[0042] Based on the clustering feature set, the training differential feature matrix is ​​optimized based on the in-class and out-of-class representations of the clustered features to obtain the training optimized differential feature matrix.

[0043] The embodiment of the present disclosure also provides a modular steel structure connection system based on a BIM model, which includes:

[0044] A connection scheme creation module, used for creating a first connection scheme of a modular steel structure and a second connection scheme of a modular steel structure in a BIM model;

[0045] A finite element analysis module, used to obtain stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure by using finite element analysis to obtain a first stress distribution diagram and a second stress distribution diagram;

[0046] A local stress pattern characteristic analysis module, used for performing local stress pattern characteristic analysis on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors;

[0047] A matrix construction module, used to construct a stress distribution adjacency characteristic matrix of the sequence of the first local stress distribution characteristic vectors and to construct a stress distribution adjacency characteristic matrix of the sequence of the second local stress distribution characteristic vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix;

[0048] A feature extraction module, used for performing feature extraction on the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively through a stress distribution topology feature extractor based on a deep neural network model to obtain a first stress distribution adjacency topology feature matrix and a second stress distribution adjacency topology feature matrix;

[0049] a feature interaction analysis module, configured to perform a graph-based local stress distribution feature interaction analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix to obtain a first solution global stress distribution feature matrix, and perform a graph-based local stress distribution feature interaction analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix to obtain a second solution global stress distribution feature matrix; and

[0050] The feature comparison module is used to compare the features of the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0051] According to the BIM model-based modular steel structure connection method and system of the embodiment of the present disclosure, different connection schemes of modular steel structures are created in the BIM model, and a data processing and analysis algorithm based on artificial intelligence is introduced in the back end to analyze the stress distribution of different connection schemes, thereby achieving more accurate connection scheme recommendations. In this way, the stress distribution in different modular steel structure connection schemes can be intelligently evaluated to automatically recommend steel structure connection schemes, which not only improves the application intelligence of BIM, but also provides decision support for the connection of modular steel structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following briefly introduces the drawings of the embodiments of the present disclosure. Obviously, the drawings described below only relate to some embodiments of the present disclosure, and are not intended to limit the present disclosure.

[0053] Figure 1 A schematic diagram of an application architecture of a modular steel structure connection method based on a BIM model in an embodiment of the present disclosure is shown;

[0054] Figure 2 A flow chart of a modular steel structure connection method based on a BIM model in an embodiment of the present disclosure is shown;

[0055] Figure 3 A flowchart of sub-step S530 of the modular steel structure connection method based on the BIM model in an embodiment of the present disclosure is shown;

[0056] Figure 4A flowchart of sub-step S570 of the modular steel structure connection method based on the BIM model in an embodiment of the present disclosure is shown;

[0057] Figure 5 A structural schematic diagram of a modular steel structure connection system based on a BIM model in an embodiment of the present disclosure is shown;

[0058] Figure 6 An application scenario diagram of a modular steel structure connection method based on a BIM model in an embodiment of the present disclosure is shown;

[0059] Figure 7 A schematic cross-sectional view of a mesh support in a modular steel structure connection solution according to an embodiment of the present disclosure is shown; and

[0060] Figure 8 A schematic diagram of a storage medium according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present disclosure, rather than all embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present disclosure.

[0062] The terms used in this specification are those general terms currently widely used in the art in consideration of the functions of the present disclosure, but these terms may vary according to the intention of a person of ordinary skill in the art, precedents, or new technologies in the art. In addition, specific terms may be selected by the applicant, and in this case, their detailed meanings will be described in the detailed description of the present disclosure. Therefore, the terms used in the specification should not be understood as simple names, but rather as an overall description based on the meaning of the terms and the present disclosure.

[0063] Although the present disclosure makes various references to certain modules in the system according to embodiments of the present disclosure, any number of different modules can be used and run on a user terminal and / or server. The modules are illustrative only, and different aspects of the systems and methods can use different modules.

[0064] Flowcharts are used in this disclosure to illustrate the operations performed by the system according to the embodiments of the present disclosure. It should be understood that the preceding or following operations are not necessarily performed precisely in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0065] Figure 1A schematic diagram of the application architecture of a modular steel structure connection method based on a BIM model in an embodiment of the present disclosure is shown, including a server 100 and a terminal device 200.

[0066] The terminal device 200 and the server 100 can be connected via the Internet to achieve mutual communication. Optionally, the above-mentioned Internet uses standard communication technology and / or protocol. The Internet is usually the Internet, but it can also be any network, including but not limited to any combination of a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or a virtual private network. In some embodiments, the data exchanged through the network is represented by technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above-mentioned data communication technologies.

[0067] The server 100 can provide various network services for the terminal device 200, wherein the server 100 can be a single server, a server cluster consisting of several servers, or a cloud computing center. Specifically, the server 100 may include a processor 110 (Center Processing Unit, CPU), a memory 120, an input device 130, and an output device 140, etc. The input device 130 may include a keyboard, a mouse, a touch screen, etc., and the output device 140 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.

[0068] The memory 120 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 110 with program instructions and data stored in the memory 120. In the embodiment of the present disclosure, the memory 120 may be used to store the program of the modular steel structure connection method based on the BIM model in the embodiment of the present disclosure.

[0069] The processor 110 calls the program instructions stored in the memory 120, and the processor 110 is used to execute the steps of any one of the modular steel structure connection methods based on the BIM model in the embodiments of the present disclosure according to the obtained program instructions.

[0070] In addition, the application architecture diagram in the embodiment of the present disclosure is intended to more clearly illustrate the technical solution in the embodiment of the present disclosure, and does not constitute a limitation on the technical solution provided in the embodiment of the present disclosure. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present disclosure is also applicable to similar problems.

[0071] The following is a non-restrictive description of the modular steel structure connection method based on the BIM model provided according to at least one embodiment of the present disclosure through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present disclosure.

[0072] Building Information Modeling (BIM) is a digital technology used to create and manage virtual representations of construction projects. BIM models contain a large amount of information about the building's geometry, materials, and systems. In the application of Building Information Modeling (BIM) technology, the connection method of modular steel structures is the key to improving construction efficiency and quality. By utilizing BIM models, complex structural analysis can be performed in the design stage, the steel structure connection scheme can be optimized, and the safety and reliability of the structure can be ensured.

[0073] Based on this, in the technical solution of the present application, a modular steel structure connection method based on the BIM model is proposed, which can better analyze and optimize the steel structure connection scheme by introducing an intelligent evaluation method in the BIM model, thereby improving the intelligence and efficiency of the building design. Specifically, the technical concept of the present application is to create different connection schemes for modular steel structures in the BIM model, and introduce artificial intelligence-based data processing and analysis algorithms in the back end to perform stress distribution analysis of different connection schemes, thereby achieving more accurate connection scheme recommendations. In this way, the stress distribution in different modular steel structure connection schemes can be intelligently evaluated to automatically recommend steel structure connection schemes, which not only improves the application intelligence of BIM, but also provides decision support for the connection of modular steel structures.

[0074] Figure 2 The flowchart of the modular steel structure connection method based on the BIM model in the embodiment of the present disclosure is shown. For example, the modular steel structure connection method based on the BIM model can be executed by a server, which can be Figure 1 The server 100 shown in FIG. Figure 2 As shown, according to the embodiment of the present disclosure, the modular steel structure connection method based on the BIM model includes the following steps: S510, creating a first connection scheme of a modular steel structure and a second connection scheme of a modular steel structure in a BIM model; S520, using finite element analysis to obtain stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain a first stress distribution diagram and a second stress distribution diagram; S530, performing local stress mode characteristic analysis on the first stress distribution diagram and the second stress distribution diagram to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors; S540, constructing a stress distribution adjacency characteristic matrix of the sequence of the first local stress distribution characteristic vectors and constructing a stress distribution adjacency characteristic matrix of the sequence of the second local stress distribution characteristic vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix; S550, by using deep The stress distribution topological feature extractor of the neural network model extracts features from the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively to obtain a first stress distribution adjacency topological feature matrix and a second stress distribution adjacency topological feature matrix; S560, performs a graph-based local stress distribution feature interactive analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topological feature matrix to obtain a first scheme global stress distribution feature matrix, and performs a graph-based local stress distribution feature interactive analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topological feature matrix to obtain a second scheme global stress distribution feature matrix; and, S570, performs a feature comparison on the first scheme global stress distribution feature matrix and the second scheme global stress distribution feature matrix to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0075] Specifically, in the technical solution of the present application, first, a modular steel structure first connection scheme and a modular steel structure second connection scheme are created in the BIM model. Next, considering that finite element analysis is a commonly used engineering analysis method, the stress distribution of the structure can be studied by numerical simulation. It takes into account the nonlinear behavior, geometry and boundary conditions of the material, thereby providing accurate stress distribution results. This enables the stress distribution, deformation and bearing capacity of the modular steel structure connection to be evaluated. Combining finite element analysis and BIM technology, the feasibility and performance of the steel structure connection scheme can be more accurately evaluated. Based on this, in the technical solution of the present application, finite element analysis is further used to obtain the stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain the first stress distribution diagram and the second stress distribution diagram.

[0076] It should be understood that since the stress distribution map is a complex data structure, it contains a large amount of information. If the feature extraction is performed directly on the entire stress distribution map, it is difficult to capture the more detailed stress distribution feature information between the modular steel structures in the connection scheme. Therefore, in order to effectively extract the detailed features of the key local areas of the modular steel structure connection in the two connection schemes, help identify the patterns and laws in the stress distribution, and thus compare the connection schemes of the two, in the technical solution of the present application, it is necessary to perform image block processing on the first stress distribution map and the second stress distribution map to obtain a sequence of the first local stress distribution map and a sequence of the second local stress distribution map.

[0077] Then, the sequence of the first local stress distribution map and the sequence of the second local stress distribution map are respectively subjected to feature mining in a stress local pattern feature extractor based on a convolutional neural network model to extract the stress distribution local pattern feature information about the modular steel structure connection in the first connection scheme and the second connection scheme, respectively, thereby obtaining a sequence of first local stress distribution feature vectors and a sequence of second local stress distribution feature vectors. In this way, the local pattern of stress distribution in the stress distribution map can be learned, so as to better understand the stress distribution of the connection scheme and help identify the stress distribution patterns and laws in different connection schemes.

[0078] Accordingly, if Figure 3As shown, in step S530, local stress pattern feature analysis is performed on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution feature vectors and a sequence of second local stress distribution feature vectors, including: S531, image block processing is performed on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution maps and a sequence of second local stress distribution maps; and, S532, the sequence of the first local stress distribution map and the sequence of the second local stress distribution map are respectively passed through a stress local pattern feature extractor based on a convolutional neural network model to obtain a sequence of the first local stress distribution feature vectors and a sequence of the second local stress distribution feature vectors.

[0079] Next, considering that there is a certain correlation and mutual influence between the local pattern features of stress distribution in each local area in the stress distribution diagram of different schemes, in order to establish this correlation relationship, so as to capture and analyze the global features of the stress distribution diagram, it is necessary to construct an adjacency feature matrix to help retain the relationship between these local features, so as to more comprehensively describe the stress distribution of the two connection schemes. Based on this, in the technical solution of the present application, a stress distribution adjacency feature matrix of the sequence of the first local stress distribution feature vectors is further constructed and a stress distribution adjacency feature matrix of the sequence of the second local stress distribution feature vectors is constructed to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix.

[0080] Furthermore, the first stress distribution adjacency matrix and the second stress distribution adjacency matrix are respectively subjected to feature mining in a stress distribution topological feature extractor based on a convolutional neural network model to respectively extract the adjacency topological association feature information between each stress distribution local area of ​​the stress distribution diagram in the first connection scheme and the second connection scheme, thereby obtaining the first stress distribution adjacency topological feature matrix and the second stress distribution adjacency topological feature matrix.

[0081] Correspondingly, in step S550, the deep neural network model is a convolutional neural network model, that is, the stress distribution topological feature extractor based on the deep neural network model is a stress distribution topological feature extractor based on the convolutional neural network model.

[0082] It should be understood that, with each first local stress distribution feature vector in the sequence of the first local stress distribution feature vector as the feature representation of the node, and with the first stress distribution adjacency topology feature matrix as the feature representation of the edge between nodes, the first global stress distribution feature matrix obtained by two-dimensional arrangement of the plurality of first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix are passed through a local stress distribution feature interactor based on a graph convolutional neural network model to obtain a first-scheme global stress distribution feature matrix. Specifically, the local stress distribution feature interactor based on the graph convolutional neural network model performs graph structure data encoding on the first global stress distribution feature matrix and the first stress distribution adjacency topology feature matrix through learnable neural network parameters to obtain the first-scheme global stress distribution feature matrix containing irregular stress distribution adjacency topology features and feature information of each local stress distribution pattern. Similarly, the sequence of the second local stress distribution feature vector and the second stress distribution adjacency topology feature matrix are also passed through the local stress distribution feature interactor based on the graph convolutional neural network model to obtain a second-scheme global stress distribution feature matrix. It should be understood that the stress distribution adjacency topology feature matrix can be used to transmit information in a graph structure and learn the association between local features. That is to say, by combining the sequence of local stress distribution eigenvectors and the adjacency topological feature matrix, the relationship between the characteristics of different local stress distribution patterns in the connection scheme and their positions and functions in the overall structure can be comprehensively considered, which helps to more comprehensively understand the stress distribution of the connection scheme and helps to more accurately evaluate the performance and reliability of the connection scheme.

[0083] Correspondingly, in step S560, the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a first-scheme global stress distribution feature matrix, and the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a second-scheme global stress distribution feature matrix, including: respectively passing the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix, and the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix through a local stress distribution feature interactor based on a graph convolutional neural network model to obtain the first-scheme global stress distribution feature matrix and the second-scheme global stress distribution feature matrix.

[0084] Furthermore, in order to be able to more intuitively compare the differences in stress distribution between the two connection schemes, in the technical solution of the present application, a differential characteristic matrix between the global stress distribution characteristic matrix of the first scheme and the global stress distribution characteristic matrix of the second scheme is further calculated. The differential characteristic matrix can help identify the global stress distribution differences and key differences between the two schemes, that is, which areas or parts have obvious stress distribution differences between the two schemes, which helps to compare, analyze and optimize the stress distribution of different connection schemes, thereby guiding design improvements and decision making.

[0085] Then, the differential feature matrix is ​​passed through a classifier-based scheme evaluator to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure. In other words, the stress distribution difference characteristics in the first connection scheme and the second connection scheme are used for classification processing to evaluate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, thereby achieving a more accurate connection scheme recommendation. In this way, the stress distribution in different modular steel structure connection schemes can be intelligently evaluated to automatically recommend connection schemes, which not only improves the application intelligence of BIM, but also provides decision support for the connection of modular steel structures.

[0086] Accordingly, if Figure 4 As shown, in step S570, a feature comparison is performed on the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including: S571, calculating the differential feature matrix between the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme; and, S572, passing the differential feature matrix through a classifier-based scheme evaluator to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0087] Among them, in step S572, the differential feature matrix is ​​passed through a classifier-based solution evaluator to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including: expanding the differential feature matrix into a classification feature vector according to a row vector or a column vector; using the fully connected layer of the classifier-based solution evaluator to fully connect encode the classification feature vector to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into the Softmax classification function of the classifier-based solution evaluator to obtain the evaluation result.

[0088] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and inefficient. Commonly used multi-classification methods include the Softmax classification function.

[0089] Furthermore, in the technical solution of the present application, the modular steel structure connection method based on the BIM model also includes a training step: for training the stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topological feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based solution evaluator.

[0090] In one example, the training step includes: obtaining training data, the training data including a first training connection scheme of a modular steel structure and a second training connection scheme of a modular steel structure created in a BIM model; using finite element analysis to obtain stress distribution maps of the first training connection scheme of the modular steel structure and the second training connection scheme of the modular steel structure to obtain a first training stress distribution map and a second training stress distribution map; performing image block processing on the first training stress distribution map and the second training stress distribution map to obtain a sequence of first training local stress distribution maps and a sequence of second training local stress distribution maps; passing the sequence of the first training local stress distribution map and the sequence of the second training local stress distribution map through the stress local pattern feature extractor based on the convolutional neural network model respectively to obtain a sequence of the first training local stress distribution feature vector and a sequence of the second training local stress distribution feature vector; constructing a stress distribution adjacency feature matrix of the sequence of the first training local stress distribution feature vector and constructing a stress distribution adjacency feature matrix of the sequence of the second training local stress distribution feature vector to obtain a first training stress distribution adjacency matrix and a second training stress distribution adjacency matrix; respectively performing image block processing on the first training stress distribution map and the second training stress distribution map through the stress local pattern feature extractor based on the convolutional neural network model The first training local stress distribution adjacency matrix and the second training stress distribution adjacency matrix are subjected to feature extraction to obtain a first training stress distribution adjacency topological feature matrix and a second training stress distribution adjacency topological feature matrix; the first training local stress distribution feature vector sequence and the first training stress distribution adjacency topological feature matrix, as well as the second training local stress distribution feature vector sequence and the second training stress distribution adjacency topological feature matrix are respectively passed through the local stress distribution feature interactor based on the graph convolutional neural network model to obtain the first training scheme global stress distribution feature matrix and the second training scheme global stress distribution feature matrix; the training differential feature matrix between the first training scheme global stress distribution feature matrix and the second training scheme global stress distribution feature matrix is ​​calculated; the training differential feature matrix is ​​optimized to obtain a training optimized differential feature matrix; the training optimized differential feature matrix is ​​passed through the classifier-based solution evaluator to obtain a classification loss function value; and the stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topological feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based solution evaluator are trained based on the classification loss function value.

[0091] In the technical solution of the present application, the global stress distribution feature matrix of the first training scheme and the global stress distribution feature matrix of the second training scheme respectively represent the stress distribution image semantic association features determined based on graph convolution coding of the stress distribution diagrams of the first training connection scheme of the modular steel structure and the second training connection scheme of the modular steel structure. However, considering that there are source domain distribution differences between the first training stress distribution diagram and the second training stress distribution diagram in the image source domain, there is a local feature distribution alignment deviation between the global stress distribution feature matrix of the first training scheme and the global stress distribution feature matrix of the second training scheme, which makes the obtained training differential feature matrix have local feature distribution structural collapse or local feature distribution discreteness when calculating the training differential feature matrix between the global stress distribution feature matrix of the first training scheme and the global stress distribution feature matrix of the second training scheme.

[0092] In this way, when the training differential feature matrix is ​​classified as a whole through the classifier-based solution evaluator, the discreteness of the local feature distribution of the training differential feature matrix or the structural collapse of the local feature distribution will make it difficult to converge to the predetermined class probability when performing class regression through the classifier-based solution evaluator, thereby affecting the training speed of the classifier-based solution evaluator and the accuracy of the final classification result.

[0093] Based on this, in the technical solution of the present application, the training differential feature matrix is ​​optimized by the following optimization formula to obtain a training optimized differential feature matrix; the optimization formula is:

[0094]

[0095] in, is the training differential feature matrix, is the scale of the training differential feature matrix, the scale of the training differential feature matrix is ​​the width multiplied by the height of the training differential feature matrix, It means point multiplication by position. Indicates adding by position point. It means to subtract by position point. To balance the hyperparameters, represents the Frobenius norm of the training differential feature matrix, Indicates the calculation of square root, is the training optimized differential feature matrix.

[0096] For the differential feature matrix as a feature set, the semantic representation of the change in units of the eigenvalues ​​of the training differential feature matrix is ​​used. In order to dynamically aggregate the semantic set composed of different change semantics of the training differential feature matrix without ignoring the individual semantic change information, the individual features of the training differential feature matrix and the aggregated scale of the differential feature matrix are expressed as the feature full amplitude and half amplitude, and the low-rank negative correlation of different dimensions of the overall semantics of the feature set of the training differential feature matrix is ​​used as the phase and scaled to dynamically adjust the change relationship of the semantic content of the training differential feature matrix, thereby improving the overall semantic information expression aggregation of the feature set of the training differential feature matrix, and improving the classification convergence efficiency of the training differential feature matrix when it is classified by the classifier-based scheme evaluator. In this way, the stress distribution in different modular steel structure connection schemes can be intelligently evaluated to automatically recommend connection schemes, which not only improves the application intelligence of BIM, but also provides decision support for the connection of modular steel structures.

[0097] Based on the above embodiments, see Figure 5As shown, it is a structural schematic diagram of a modular steel structure connection system 800 based on a BIM model in an embodiment of the present disclosure. The modular steel structure connection system 800 based on a BIM model includes: a connection scheme creation module 810, which is used to create a first connection scheme of a modular steel structure and a second connection scheme of a modular steel structure in a BIM model; a finite element analysis module 820, which is used to use finite element analysis to obtain the stress distribution diagram of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain a first stress distribution diagram and a second stress distribution diagram; a local stress pattern feature analysis module 830, which is used to perform a local stress pattern feature analysis on the first stress distribution diagram and the second stress distribution diagram to obtain a sequence of first local stress distribution feature vectors and a sequence of second local stress distribution feature vectors; a matrix construction module 840, which is used to construct a stress distribution adjacency feature matrix of the sequence of the first local stress distribution feature vectors and a stress distribution adjacency feature matrix of the sequence of the second local stress distribution feature vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix; a feature extraction module 85 0, used for performing feature extraction on the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively through a stress distribution topology feature extractor based on a deep neural network model to obtain a first stress distribution adjacency topology feature matrix and a second stress distribution adjacency topology feature matrix; a feature interaction analysis module 860, used for performing a graph-based local stress distribution feature interaction analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix to obtain a first scheme global stress distribution feature matrix, and performing a graph-based local stress distribution feature interaction analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix to obtain a second scheme global stress distribution feature matrix; and a feature comparison module 870, used for performing feature comparison on the first scheme global stress distribution feature matrix and the second scheme global stress distribution feature matrix to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0098] Here, those skilled in the art can understand that the specific functions and operations of each module in the above-mentioned modular steel structure connection system 800 based on the BIM model have been referred to above. Figures 2 to 4 The description of the modular steel structure connection method based on the BIM model has been introduced in detail, and therefore, its repeated description will be omitted.

[0099] Figure 6 FIG. 1 is an application scenario diagram of a modular steel structure connection method based on a BIM model according to an embodiment of the present disclosure. Figure 6 As shown, in this application scenario, first, a modular steel structure first connection scheme is created in the BIM model (for example, Figure 6 D1) and the modular steel structure second connection scheme (e.g. Figure 6 Then, the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure are input to a server (for example, Figure 6 S) as shown in , wherein the server is capable of using the modular steel structure connection algorithm based on the BIM model to process the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain an evaluation result indicating whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

[0100] Further, Figure 7 FIG. 2 shows a schematic cross-sectional view of a mesh support in a modular steel structure connection solution according to an embodiment of the present disclosure. Figure 7 As shown, the mesh base support includes a connecting portion 10, a rib plate 20 and a bottom plate 40 for connecting other components. The connecting portion 10, the rib plate 20 and the bottom plate 40 are interconnected and fixedly connected to the original steel column top plate 60 through at least two anchor bolts 30, wherein a transition plate 50 is also provided between the bottom plate 40 and the original steel column top plate 60 to improve the connection performance.

[0101] Based on the above embodiments, another exemplary embodiment of an electronic device is also provided in the embodiments of the present disclosure. In some possible implementations, the electronic device in the embodiments of the present disclosure may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the modular steel structure connection method based on the BIM model in the above embodiments can be implemented when the processor executes the program.

[0102] For example, the electronic device is disclosed in the present invention Figure 1 Taking the server 100 in the example as an example, the processor in the electronic device is the processor 110 in the server 100, and the memory in the electronic device is the memory 120 in the server 100.

[0103] An embodiment of the present disclosure also provides a computer-readable storage medium. Figure 8 FIG. 1 is a schematic diagram of a computer-readable storage medium 1000 according to an embodiment of the present disclosure. Figure 8As shown, the computer-readable storage medium 1000 stores computer-executable instructions 1001. When the computer-executable instructions 1001 are executed by the processor, the modular steel structure connection method based on the BIM model according to the embodiment of the present disclosure described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc.

[0104] The embodiment of the present disclosure also provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the modular steel structure connection method based on the BIM model according to the embodiment of the present disclosure.

[0105] Those skilled in the art will appreciate that the contents disclosed in this disclosure may be subject to various modifications and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0106] In addition, although the present disclosure makes various references to certain units in the system according to embodiments of the present disclosure, any number of different units may be used and run on the client and / or server. The units are illustrative only, and different aspects of the systems and methods may use different units.

[0107] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of a software functional module. The present disclosure is not limited to any particular form of combination of hardware and software.

[0108] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and should not be interpreted in an idealized or highly formal sense, unless explicitly defined as such herein.

[0109] The above is an explanation of the present disclosure and should not be considered as a limitation thereof. Although several exemplary embodiments of the present disclosure have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present disclosure. Therefore, all such modifications are intended to be included within the scope of the present disclosure as defined in the claims. It should be understood that the above is an explanation of the present disclosure and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present disclosure is defined by the claims and their equivalents.

Claims

1. A modular steel structure connection method based on BIM model, characterized in that: include: Create a first connection scheme of modular steel structure and a second connection scheme of modular steel structure in the BIM model; Using finite element analysis to obtain stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure to obtain a first stress distribution diagram and a second stress distribution diagram; Performing local stress mode characteristic analysis on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors; Constructing a stress distribution adjacency characteristic matrix of the sequence of the first local stress distribution characteristic vectors and constructing a stress distribution adjacency characteristic matrix of the sequence of the second local stress distribution characteristic vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix; Performing feature extraction on the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively by a stress distribution topology feature extractor based on a deep neural network model to obtain a first stress distribution adjacency topology feature matrix and a second stress distribution adjacency topology feature matrix; Performing a graph-structure-based local stress distribution feature interactive analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix to obtain a first-solution global stress distribution feature matrix, and performing a graph-structure-based local stress distribution feature interactive analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix to obtain a second-solution global stress distribution feature matrix; as well as A feature comparison is performed on the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

2. The modular steel structure connection method based on the BIM model according to claim 1 is characterized in that: Performing local stress mode characteristic analysis on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors, including: Performing image block processing on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution maps and a sequence of second local stress distribution maps; and The sequence of the first local stress distribution maps and the sequence of the second local stress distribution maps are respectively passed through a stress local pattern feature extractor based on a convolutional neural network model to obtain a sequence of the first local stress distribution feature vectors and a sequence of the second local stress distribution feature vectors.

3. The modular steel structure connection method based on the BIM model according to claim 2 is characterized in that: The deep neural network model is a convolutional neural network model.

4. The modular steel structure connection method based on the BIM model according to claim 3 is characterized in that: The sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a first-solution global stress distribution feature matrix, and the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix are subjected to a graph-based local stress distribution feature interactive analysis to obtain a second-solution global stress distribution feature matrix, including: The sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix, as well as the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix are respectively passed through a local stress distribution feature interactor based on a graph convolutional neural network model to obtain the first scheme global stress distribution feature matrix and the second scheme global stress distribution feature matrix.

5. The modular steel structure connection method based on the BIM model according to claim 4 is characterized in that: Comparing the characteristics of the global stress distribution characteristic matrix of the first scheme and the global stress distribution characteristic matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including: Calculating a differential characteristic matrix between the global stress distribution characteristic matrix of the first solution and the global stress distribution characteristic matrix of the second solution; and The differential feature matrix is ​​passed through a classifier-based solution evaluator to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection solution of the modular steel structure is better than the second connection solution of the modular steel structure.

6. The modular steel structure connection method based on the BIM model according to claim 5 is characterized in that: The differential feature matrix is ​​passed through a scheme evaluator based on a classifier to obtain an evaluation result, and the evaluation result is used to indicate whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure, including: Expanding the differential feature matrix into classification feature vectors according to row vectors or column vectors; Performing fully connected encoding on the classification feature vector using the fully connected layer of the classifier-based solution evaluator to obtain an encoded classification feature vector; and The encoded classification feature vector is input into the Softmax classification function of the classifier-based solution evaluator to obtain the evaluation result.

7. The modular steel structure connection method based on the BIM model according to claim 6 is characterized in that: It also includes a training step: for training the stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topology feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based solution evaluator.

8. The modular steel structure connection method based on the BIM model according to claim 7 is characterized in that: The training step comprises: Acquire training data, wherein the training data includes a first training connection scheme of a modular steel structure and a second training connection scheme of a modular steel structure created in a BIM model; Using finite element analysis to obtain stress distribution diagrams of the first training connection scheme of the modular steel structure and the second training connection scheme of the modular steel structure to obtain a first training stress distribution diagram and a second training stress distribution diagram; Performing image block processing on the first training stress distribution map and the second training stress distribution map to obtain a sequence of first training local stress distribution maps and a sequence of second training local stress distribution maps; Passing the sequence of the first training local stress distribution maps and the sequence of the second training local stress distribution maps through the stress local pattern feature extractor based on the convolutional neural network model to obtain a sequence of first training local stress distribution feature vectors and a sequence of second training local stress distribution feature vectors respectively; Constructing a stress distribution adjacency feature matrix of the first training sequence of local stress distribution feature vectors and constructing a stress distribution adjacency feature matrix of the second training sequence of local stress distribution feature vectors to obtain a first training stress distribution adjacency matrix and a second training stress distribution adjacency matrix; The stress distribution topology feature extractor based on the convolutional neural network model performs feature extraction on the first training stress distribution adjacency matrix and the second training stress distribution adjacency matrix respectively to obtain a first training stress distribution adjacency topology feature matrix and a second training stress distribution adjacency topology feature matrix; The sequence of the first training local stress distribution feature vectors and the first training stress distribution adjacency topology feature matrix, as well as the sequence of the second training local stress distribution feature vectors and the second training stress distribution adjacency topology feature matrix are respectively passed through the local stress distribution feature interactor based on the graph convolutional neural network model to obtain a first training scheme global stress distribution feature matrix and a second training scheme global stress distribution feature matrix; Calculating a training difference feature matrix between the global stress distribution feature matrix of the first training scheme and the global stress distribution feature matrix of the second training scheme; Optimizing the training differential feature matrix to obtain a training optimized differential feature matrix; Passing the training optimized differential feature matrix through the classifier-based solution evaluator to obtain a classification loss function value; and Based on the classification loss function value, the stress local pattern feature extractor based on the convolutional neural network model, the stress distribution topology feature extractor based on the convolutional neural network model, the local stress distribution feature interactor based on the graph convolutional neural network model and the classifier-based solution evaluator are trained.

9. A modular steel structure connection system based on a BIM model, characterized in that: include: A connection scheme creation module, used for creating a first connection scheme of a modular steel structure and a second connection scheme of a modular steel structure in a BIM model; A finite element analysis module, used to obtain stress distribution diagrams of the first connection scheme of the modular steel structure and the second connection scheme of the modular steel structure by using finite element analysis to obtain a first stress distribution diagram and a second stress distribution diagram; A local stress pattern characteristic analysis module, used for performing local stress pattern characteristic analysis on the first stress distribution map and the second stress distribution map to obtain a sequence of first local stress distribution characteristic vectors and a sequence of second local stress distribution characteristic vectors; A matrix construction module, used to construct a stress distribution adjacency characteristic matrix of the sequence of the first local stress distribution characteristic vectors and to construct a stress distribution adjacency characteristic matrix of the sequence of the second local stress distribution characteristic vectors to obtain a first stress distribution adjacency matrix and a second stress distribution adjacency matrix; A feature extraction module, used for performing feature extraction on the first stress distribution adjacency matrix and the second stress distribution adjacency matrix respectively through a stress distribution topology feature extractor based on a deep neural network model to obtain a first stress distribution adjacency topology feature matrix and a second stress distribution adjacency topology feature matrix; a feature interaction analysis module, configured to perform a graph-based local stress distribution feature interaction analysis on the sequence of the first local stress distribution feature vectors and the first stress distribution adjacency topology feature matrix to obtain a first scheme global stress distribution feature matrix, and to perform a graph-based local stress distribution feature interaction analysis on the sequence of the second local stress distribution feature vectors and the second stress distribution adjacency topology feature matrix to obtain a second scheme global stress distribution feature matrix; as well as The feature comparison module is used to compare the features of the global stress distribution feature matrix of the first scheme and the global stress distribution feature matrix of the second scheme to determine whether the first connection scheme of the modular steel structure is better than the second connection scheme of the modular steel structure.

Images