Building structure quality risk prediction method, device, equipment, medium and product

Abstract

The invention discloses a building structure quality risk prediction method and device, equipment, a medium and a product. The method comprises the steps that structure extraction is conducted on a pre-constructed target building information model, multiple pieces of target building structure data are obtained, and the target building structure data at least comprise building foundation structure data, building main body structure data and building floor structure data; inputting the target building structure data into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result, the target prediction model is obtained by training according to a plurality of first building structure data samples and second building structure data samples obtained after the feature spaces of the first building structure data samples are converted. According to the embodiment of the invention, the target prediction model can maintain high prediction accuracy and robustness when processing building structure data of different types and complexities.

Description

Technical Field

[0001] The present application belongs to the field of building safety technology, and in particular, relates to a method, device, equipment, medium and product for predicting building structure quality risks. Background Art

[0002] In the field of construction engineering, the risk prediction of building structure quality is of vital importance, which involves the assessment of the stability, safety and durability of the building structure. With the rapid development of the construction industry, the complexity and diversity of building structures are increasing, which puts higher requirements on the accuracy of the risk prediction of building structure quality.

[0003] Existing building structure quality risk prediction models are mainly based on traditional statistical methods and machine learning algorithms. These methods perform well when processing simple and standardized building structure data, but when faced with complex and diverse building structure data, it is often difficult to accurately capture the building structure characteristics and the complex relationships between them, resulting in deviations in the prediction results. Therefore, when facing new types of building structure data or building structure data with high structural complexity, the prediction performance of the models in the relevant technologies often decreases significantly, showing poor robustness. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment, medium and product for predicting building structure quality risks, which are used to at least solve the problems of insufficient prediction accuracy and poor robustness in the related art when processing complex and diverse building structure data.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting building structure quality risks, comprising:

[0006] Performing structural extraction on a pre-constructed target building information model to obtain a plurality of target building structure data, wherein the target building structure data at least includes building foundation structure data, building main structure data, and building floor structure data;

[0007] The target building structure data is input into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result, wherein the target prediction model is trained based on multiple first building structure data samples and second building structure data samples after transforming the feature space of the first building structure data samples.

[0008] In a second aspect, an embodiment of the present application provides a device for predicting building structure quality risks, the device comprising:

[0009] An extraction module is used to perform structural extraction on a pre-built target building information model to obtain a number of target building structure data, wherein the target building structure data at least includes building foundation structure data, building main structure data, and building floor structure data;

[0010] An input module is used to input the target building structure data into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result. The target prediction model is trained based on multiple first building structure data samples and second building structure data samples after transforming the feature space of the first building structure data samples.

[0011] In a third aspect, an embodiment of the present application provides an electronic device comprising: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the steps of the building structure quality risk prediction method as described in any one of the embodiments of the first aspect are implemented.

[0012] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the steps of the method for predicting building structure quality risks as described in any one of the embodiments of the first aspect are implemented.

[0013] In a fifth aspect, an embodiment of the present application provides a computer program product, which is stored in a storage medium and is executed by at least one processor to implement the steps of the building structure quality risk prediction method provided in the first aspect of the embodiment of the present application.

[0014] The building structure quality risk prediction method, device, equipment, medium and product of the embodiments of the present application iteratively optimize the preset model based on the sample building structure data after feature space transformation, and continuously adjust the model parameters so that the trained target prediction model can maintain high prediction accuracy and robustness when processing building structure data of different types and complexities. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0016] Figure 1 It is a flow chart of a method for predicting building structure quality risk provided in an embodiment of the present application;

[0017] Figure 2 It is a flowchart of a method for training a target prediction model provided in an embodiment of the present application;

[0018] Figure 3 It is a structural schematic diagram of a building structure quality risk prediction device provided in an embodiment of the present application;

[0019] Figure 4 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application.

[0020] Reference numerals:

[0021] Building structure quality risk prediction device 300, extraction module 301, input module 302,

[0022] Electronic device 400 , processor 401 , memory 402 , communication interface 403 , bus 410 . DETAILED DESCRIPTION

[0023] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0024] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0025] In the field of construction engineering, the risk prediction of building structure quality is of vital importance, which involves the assessment of the stability, safety and durability of the building structure. With the rapid development of the construction industry, the complexity and diversity of building structures are increasing, which puts higher requirements on the accuracy of the risk prediction of building structure quality.

[0026] Existing building structure quality risk prediction models are mainly based on traditional statistical methods and machine learning algorithms. These methods perform well when processing simple and standardized building structure data, but when faced with complex and diverse building structure data, it is often difficult to accurately capture the building structure characteristics and the complex relationships between them, resulting in deviations in the prediction results.

[0027] Specifically, when dealing with complex and diverse building structure data, the building structure quality risk prediction models in related technologies often find it difficult to accurately capture structural features and the complex relationships between them. This is because the representation of data of different building types and structural complexity in the feature space varies greatly, and existing models are difficult to adapt effectively. Due to the difficulty in accurately capturing complex structural features, related technologies often have large errors when predicting building structure quality risks, resulting in inaccurate prediction results. Therefore, when faced with new types of building structure data or building structure data with high structural complexity, the prediction performance of the models in related technologies often decreases significantly, showing poor robustness.

[0028] In order to solve the problems of related technologies, the embodiments of the present application provide a method, device, equipment, medium and product for predicting the quality risks of building structures.

[0029] The following is a detailed description of the building structure quality risk prediction method provided by the embodiment of the present application through specific embodiments and their application scenarios in conjunction with the accompanying drawings.

[0030] Figure 1 FIG. 1 is a flow chart showing a method for predicting building structure quality risk according to an embodiment of the present application. Figure 1 As shown, the building structure quality risk prediction method may specifically include the following steps:

[0031] S101, performing structural extraction on a pre-constructed target building information model to obtain a plurality of target building structural data, wherein the target building structural data at least includes building foundation structural data, building main structural data, and building floor structural data;

[0032] S102. Input the target building structure data into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result, wherein the target prediction model is trained based on a plurality of first building structure data samples and second building structure data samples after transforming the feature space of the first building structure data samples.

[0033] Therefore, based on the sample building structure data after feature space transformation, the preset model is iteratively optimized. By continuously adjusting the model parameters, the trained target prediction model can maintain high prediction accuracy and robustness when processing building structure data of different types and complexities.

[0034] It should be noted that the embodiment of the present application includes multiple target building information models corresponding to multiple different building projects. Taking a new high-rise residential building project as an example, its target building information model may include overall building planning information, building functional layout information, building structure design parameters, etc. Then, for the target building structure data in S101, at least one building structure data is generated by structural extraction from the target building information model.

[0035] In some embodiments, in S101, during the structure extraction process, information directly related to the building structure is screened out from the target building information model. Specifically, for high-rise residential buildings, building infrastructure information is extracted, including but not limited to the type of foundation (such as pile foundation), the size of the foundation (such as pile diameter, pile length, etc.), and the design value of the bearing capacity of the foundation; for the main structure of the building, the distribution of columns, the size of columns (such as diameter, height, etc.), the arrangement of beams, the size of beams (such as width, height, etc.), and the building materials used for beams and columns (such as the strength grade of concrete, the model of steel, etc.) are extracted; and building floor structure information, such as the thickness of the floor slab, the structural form of the stairs (such as plate stairs or beam stairs), etc.

[0036] Optionally, the target building information model can also be the target building information model corresponding to the commercial complex construction project. For commercial complexes, due to their complex structure, the large-span structural requirements of commercial spaces need to be considered during the structural extraction process. Therefore, the target building structure data includes not only the building infrastructure data, the building main structure data, and the building floor structure data, but also information related to special structures such as large-span beams and arches, such as their structural form, force characteristics, material properties, etc. In addition, since there may be a large number of changes in passenger flow loads in commercial complexes, the target building structure data also includes information related to the structure of personnel evacuation passages, such as but not limited to: the width of the evacuation stairs, the structural stability of the evacuation passages, etc.

[0037] In this way, the building structure data extracted from a plurality of different target building information models are used as target building structure data to be input into the target prediction model for building structure quality risk prediction.

[0038] It should be noted that, before S102, a target prediction model is constructed and trained to output building structure quality risk prediction results using the target prediction model, thereby providing reliable protection for the safety and durability of the construction project.

[0039] Specifically, the training process of the target prediction model provided in the embodiment of the present application can refer to Figure 2 .like Figure 2As shown, a target prediction model training method provided in an embodiment of the present application may include the following steps: S201 to S204.

[0040] S201, obtaining a plurality of first building structure data samples;

[0041] S202, using the transformation iterative learning sub-instruction to perform feature space transformation on the first building structure data sample to obtain a second building structure data sample after the feature space transformation;

[0042] S203, generating, according to the first building structure data sample and the second building structure data sample, a training cost result when a model iteration instruction for running building structure quality risk prediction corresponding to a preset model is executed;

[0043] S204, adjusting the weight parameters of the preset model according to the training cost result, and training the adjusted preset model until a preset training stop condition is met, thereby obtaining a trained target prediction model.

[0044] It should be noted that the preset model in the embodiment of the present application can be applied to the risk prediction of building structure quality of various types of buildings, including but not limited to residential buildings, commercial buildings, industrial plants, etc. Then, the sample data used for model training includes a large number of building structure data samples corresponding to various types of buildings, which is not specifically limited in this embodiment.

[0045] Optionally, the first building structure data sample includes at least structural mechanics parameters.

[0046] In some embodiments, in S202, when the initial state of the first building structure data sample is the first feature space, an association relationship between the structural mechanics parameters and pre-set non-structural factor parameters is established to obtain a second building structure data sample after the feature space is transformed, and the state of the second building structure data sample is the second feature space. The second building structure data sample includes structural mechanics parameters, non-structural factor parameters and characteristics of the association relationship between the building structure and the non-structural factors; when the initial state of the first building structure data sample is the second feature space, the association relationship between the building structure and the non-structural factors is simplified to obtain a second building structure data sample after the feature space is transformed, and the state of the second building structure data sample is the first feature space.

[0047] That is to say, for each first building structure data sample, the model iteration instruction of feature space transformation must be executed. Taking the first building structure data sample of a multi-story residential building as an example, it at least includes building infrastructure data (such as foundation type, depth, etc.), building main structure data (such as beam and column size, material, etc.), building floor structure data (such as floor thickness, staircase structure, etc.). It should be noted that these data are in a specific feature space in the initial state, which is defined as the first feature space or the second feature space in this embodiment.

[0048] Specifically, the first building structure data sample whose initial state is the first feature space is referred to as the first sample building structure data, and the first building structure data sample whose initial state is the second feature space is referred to as the second sample building structure data, and the specific implementation methods of transforming the feature spaces of these two types of data samples are described in detail.

[0049] For the first example of building structure data, for example, in the sample data of a multi-story residential building, the basic structure data may be represented in a specific format and value range, which is a manifestation of the first feature space. The model iteration instruction of the first feature space transformation is run from the first feature space to the second feature space. Among them, the second feature space is determined through research and analysis, and is an ideal feature space for the model to adapt when performing risk prediction for parameter iteration optimization of the building structure quality risk prediction model.

[0050] In this process, the foundation type in the infrastructure data needs to be converted. In the first feature space, the foundation type is usually represented by traditional classification methods, such as simple classifications such as shallow foundation and deep foundation. When converting to the second feature space, it is necessary to refine it into more precise types, including but not limited to raft foundation, pile foundation, etc., and establish an association with relevant geological parameters. At the same time, for the size data of beams and columns, in the first feature space, it is represented by the actual length and width values. When converting to the second feature space, it is converted into a relative value related to the overall structural proportion of the building, and a new connection is established with factors such as the load borne by the building.

[0051] For the second sample building structure data, taking a commercial building as an example, its initial feature space is the second feature space, and it is necessary to run the model iteration instruction of the second feature space transformation from the second feature space to the first feature space. The structural data of the commercial building is relatively complex. In the second feature space, its structural data is in a highly optimized and complexly associated form. For example, its column grid structure data may be associated with multiple factors such as the functional zoning of the building and the distribution of pedestrian flow. When converting to the first feature space, these complex associations need to be simplified, and the column grid structure data needs to be converted into a more basic representation. Specifically, it can be represented according to simple geometric shapes and sizes, while weakening the association with other non-structural factors.

[0052] It can be understood that in the execution process of the model iteration instruction of the entire feature space transformation, for different sample building structure data, whether it is the transformation from the first feature space to the second feature space, or the transformation from the second feature space to the first feature space, it is necessary to follow specific algorithms and rules, namely, the transformation iteration learning sub-instructions. Specifically, the feature space transformation algorithm is determined based on the principles of building structure engineering, the pre-set quality risk assessment requirements, and the model optimization target information.

[0053] In the specific implementation, during the conversion process, factors such as the mechanical properties of building materials and the stability of the structure need to be considered. For different building structure parts, such as beams and columns in the frame structure, when the feature space is transformed, the conversion operation is performed according to their force characteristics and mutual relationships in the entire building structure.

[0054] In this way, the sample building structure data is transformed from the original feature space to a feature space that is more suitable for model prediction. This transformation process can capture the structural features and the complex relationships between them, so that the model can better understand the data structure. That is, the feature space can be dynamically adjusted according to the type and complexity of the building structure data, so that the building structure quality risk prediction model can better adapt to data of different types and complexities. This key innovation lies in the dynamic and adaptive nature of the feature space transformation, which significantly improves the prediction accuracy and robustness of the model.

[0055] Further, in some embodiments, in S203, the second building structure data sample after the feature space is transformed is subjected to an inverse transformation of the feature space to obtain a third building structure data sample; the deviation between the first building structure data sample and the third building structure data sample is determined; when the deviation is greater than or equal to a preset threshold, adjustment information of the transformation algorithm corresponding to the transformation iterative learning sub-instruction is determined according to the deviation; the transformation iterative learning sub-instruction is optimized according to the adjustment information; and a training cost result is generated according to the deviation and the adjustment information.

[0056] In addition, in some embodiments, multiple second building structure data samples are input into the encoding module of the preset model; the encoding module is used to encode the multiple second building structure data samples respectively to obtain multiple sample feature vectors; and the encoding module is used to cross-fuse the multiple sample feature vectors in the spatial dimension.

[0057] In the specific implementation, the internal parameters of the encoding module are set, including but not limited to the weights, bias terms, activation function types, and learning rates of the neural network layers, so as to initialize the encoding module. In addition, appropriate data structures such as tensors, lists, or matrices are prepared for the intermediate data and final results in the encoding process to efficiently store and access data.

[0058] In a specific implementation, the input data of the encoding module is a plurality of sample building structure data after feature space transformation (i.e., second building structure data samples). These data have been transformed into the second feature space through feature space transformation, and are denoted as {S1', S2', ..., S n '}, where n is the number of sample building structure data.

[0059] In specific implementation, for each sample building structure data S i 'Perform normalization processing, which may include normalization, standardization, and data enhancement, to ensure the stability and consistency of the data during the encoding process, and output the normalized data {S1'', S2'', ..., S n ''}; Then, from the normalized data {S1'', S2'', ..., S n ''} extract key features, which include structural size, material type, load-bearing capacity, etc., and output feature set {F1, F2,..., F n}, complete the preprocessing of the sample building structure data.

[0060] In specific implementation, each feature F i The first layer (fully connected layer or convolutional layer) of the encoding module is input, and the features are linearly transformed and nonlinearly activated through the weights and biases of this layer, and the preliminary encoded features {E1, E2, ..., E n}; for the initial encoded features {E1, E2, ..., E n} Perform feature fusion and output the fused feature E fused , where the specific implementation methods of feature fusion may include concatenation, weighted summation, and attention fusion mechanism; further, the fused feature E fusedThe deep network structure is input into the encoding module. These deep structures include multiple convolutional layers, pooling layers, fully connected layers or recurrent neural network layers. Through layer-by-layer processing, higher-level feature representations are gradually extracted. After each layer or a specific layer, the features are nonlinearly transformed through activation functions to enhance the expression ability of the model. At the same time, dropout, batch normalization and other technologies are used to prevent overfitting and accelerate convergence, and the deep-encoded features E are output. deep , where the activation function may include ReLU, Sigmoid or Tanh, which is not specifically limited in this embodiment; in the spatial dimension, for E deep Specifically, it can be achieved by presetting the network structure (cross attention network or graph convolution network) or using a specific fusion algorithm to capture the complex relationship and interaction between the features and output the cross-fused feature E cross ; for E cross Transform the spatial dimension. Specifically, operations such as transposition, reshaping or expansion can be performed to convert it into a form suitable for subsequent processing or decoding, and the transformed feature E is output. transformed ; E transformed The input is sent to the last layer of the encoding module (i.e., the fully connected layer or the output layer). After being processed by this layer, the final encoding representation {C1, C2, ..., C n}, where C i is the sample building structure data S i ', it should be understood that the encoding representation {C1, C2, ..., C n} is the input of the preset model and is used for subsequent building structure quality risk prediction tasks.

[0061] The above encoding process is further described in detail below using scenario examples.

[0062] The building structure quality risk prediction model is used to encode and represent the sample building structure data after multiple transformed feature spaces generated by the model iteration instructions of the feature space transformation. Taking the building structure data of the large commercial center and high-rise office building mentioned above as an example, these data are in the second feature space after the feature space transformation and serve as the input data of the model.

[0063] First, the encoding module parameters and data structure are initialized. The input feature space is transformed into multiple sample building structure data, such as the raft foundation data (thickness, reinforcement, etc.) of the foundation structure of a large commercial center, the giant column data (structural form, combination method, etc.) of the main structure, and the atrium large-span steel structure roof data (structural parameters, etc.), as well as the foundation structure pile foundation data of the high-rise office building, the main structure frame column and beam data, the floor structure floor and staircase data, etc., recorded as {S1', S2', ..., S n '}. Set internal parameters for the encoding module. For example, the weights of the neural network layer can be initialized to specific values ​​according to the scale and complexity of the building structure data. The bias term will also be set to an appropriate value. The activation function type can choose the ReLU function, and the learning rate is set to a small value (such as 0.001) to ensure the stability of training. At the same time, prepare data structures for intermediate data and final results, such as storing the basic structure data in tensor form and the main structure data in matrix form, so as to efficiently store and access data.

[0064] Furthermore, the sample building structure data is preprocessed. For the input {S1', S2', ..., S n '}, perform data normalization. Taking the raft foundation thickness data of a large commercial center as an example, its original value may be in a large range, and it is converted to the [0, 1] interval through normalization. For the concrete strength grade data of the frame columns of high-rise office buildings, standardization is performed to make them conform to the standard normal distribution. At the same time, data enhancement operations can be performed, such as slightly randomly perturbing the structural parameter data of the large-span steel structure roof of the atrium of the commercial center to increase the diversity of the data, and outputting the normalized data {S1'', S2'', ..., S n ''}. Then from {S1'', S2'', ..., S n ''}. For the raft foundation of the commercial center, its bearing capacity, connection characteristics with surrounding structures and other features are extracted; for the frame columns of the office building, its cross-sectional dimensions, steel strength and other features are extracted, and the feature set {F1, F2, ..., F n}.

[0065] Secondly, feature encoding and preliminary fusion are performed. Each feature F i Input to the first layer of the encoding module. For example, for the bearing capacity characteristics of the raft foundation of the commercial center, the weights and biases of the fully connected layer are linearly transformed, and then nonlinearly activated by the ReLU activation function to obtain the initial encoded features. The same operation is performed for the cross-sectional size characteristics of the frame columns of the office building, and the initial encoded features {E1, E2, ..., E nThen, these preliminary coded features are fused. For the structural data of commercial centers and office buildings, a weighted summation method can be used to assign different weights to different structural parts according to their importance to the overall structural risk, and the preliminary coded features can be fused into E fused .

[0066] Then, perform deep coding and feature extraction. fused Input into the deep network structure of the encoding module. Among them, the deep structure includes a convolution layer. For the structural data of the commercial center, the convolution layer will extract the structural features of different local areas, and the pooling layer will downsample these features to reduce the amount of data. For the structural data of the office building, the global features are further integrated when passing through the fully connected layer. After each layer, the ReLU activation function can be used to perform nonlinear transformation on the features. For example, for the structural features of the large-span steel structure roof of the atrium of the commercial center, its complex structural relationship can be better represented after the activation function. At the same time, the dropout technology is used to prevent overfitting, such as randomly disconnecting the neuron connection with a certain probability (such as 0.5) in the fully connected layer, and outputting the deep encoded feature E deep .

[0067] Next, feature cross-fusion and spatial dimension transformation are performed. deep For the structural data of commercial centers and office buildings, the features of different structural parts such as raft foundation and frame columns, long-span roof and floor slab are interrelated through special network structures such as cross-attention network. For example, the structural stability features of the long-span roof and the supporting capacity features of the frame columns are cross-fused to capture the complex relationship between them, and the cross-fused features E are output. cross Then for E cross Perform spatial dimension transformation, such as converting it from a three-dimensional structural feature representation to a two-dimensional form suitable for subsequent processing, and output the transformed feature E transformed .

[0068] Finally, the final generation of the encoding representation is achieved. transformed Input to the last layer of the encoding module. For the structure data of commercial centers and office buildings, the last layer is a fully connected layer. Through the processing of this layer, the final encoding representation {C1, C2, ..., C n}. Among them, C1 is used to represent the coded representation of the raft foundation structure data of the commercial center, and C2 is used to represent the coded representation of the frame column structure data of the office building.

[0069] As an optional embodiment, the coded representation can be verified and optimized, and then the optimized coded representation can be stored in an appropriate storage medium (such as memory, database or file system) so that the subsequent model can quickly access and call it, that is, in the task of predicting the quality risk of building structures, the stored coded representation is used as input data and input into the trained target prediction model, and the quality risk prediction result of the building structure is output through the prediction and decision of the target prediction model. For example, when predicting the structural quality risk of a commercial center, the model will output the quality risk prediction result of the building structure based on the coded representation of the structural data such as the raft foundation and giant column, combined with the pre-trained model parameters.

[0070] In specific implementation, the quality of the coded representations is verified by calculating indicators such as the similarity, distance or reconstruction error between the coded representations to ensure that they can accurately reflect the characteristics of the original sample building structure data; based on the verification results, the parameters of the coding module are adjusted and optimized. Specifically, the weights and bias terms can be updated through the back-propagation algorithm, or the regularization technique can be used to reduce the complexity of the model to improve the accuracy and robustness of the coded representation.

[0071] Taking the coding representation of the structural data of commercial centers and office buildings as an example, the cosine similarity between them can be calculated to ensure that the coding representation of different structural data can accurately reflect the characteristics of the original data. According to the verification results, the parameters of the coding module are adjusted through the back propagation algorithm. If it is found that the accuracy of the coding representation is insufficient, such as the deviation in the structural association between the coding representation of the raft foundation and the frame column, the weights and bias items of the neural network layer are adjusted, or the L2 regularization technology is used to reduce the complexity of the model and improve the accuracy and robustness of the coding representation.

[0072] In this way, by designing the encoding module, the sample building structure data after the feature space transformation is encoded and represented by the encoding module, and cross-fused in the spatial dimension. This feature encoding and fusion mechanism can capture the complex relationship between structural features, further enhance the model's ability to understand complex structural features, and thus improve the model's prediction performance.

[0073] Furthermore, in some embodiments, in S204, the preset model after adjustment is trained with multiple sample feature vectors as input of the preset model until a preset training stop condition is met, thereby obtaining a trained target prediction model.

[0074] In some embodiments, the building structure quality risk training results of the sample building structure data after multiple transformation feature spaces are determined based on the sample building structure features after the coded representation. Taking the commercial center and the office building as an example, for the basic structure of the commercial center, according to the bearing capacity related information in the coded representation of the raft foundation, combined with the risk assessment standard in the building structure quality risk prediction model, if the bearing capacity coded representation of the raft foundation shows that it may face overload risks in the future (possibly due to the increase in load caused by the expansion of the commercial center function), then the corresponding risk level is marked in the risk training results of the basic structure. For the main structure of the office building, according to the structural strength information in the coded representation of the frame columns and beams, if the coded representation of the frame columns and beams shows that their structural strength is insufficient considering factors such as earthquake loads, the risk situation is marked in the risk training results of the main structure.

[0075] In some embodiments, based on the building structure quality risk training results of the sample building structure data after multiple transformation feature spaces, the training cost results of the building structure quality risk prediction model when running the model iteration instruction of the building structure quality risk prediction are generated. If in the risk training results of the commercial center, it is found that the risk assessment errors of the raft foundation and the atrium large-span steel structure roof are large, such as the actual risk exists but the model does not accurately identify it or vice versa, then the training cost result will be high, indicating that the model needs more optimization in this regard. For office buildings, if the risk assessment of frame columns and floor slabs is inaccurate, it will also lead to an increase in the training cost results.

[0076] In another embodiment, the training cost result corresponding to the model iteration instruction of feature space transformation and the training cost result corresponding to the model iteration instruction of building structure quality risk prediction are determined; the training cost result corresponding to the model iteration instruction of feature space transformation and the training cost result of the model iteration instruction of building structure quality risk prediction are weighted and calculated, and the parameters of the preset model are iteratively optimized based on the calculation result until the training termination condition is met. In this way, based on the model iteration instruction of feature space transformation and the model iteration instruction of building structure quality risk prediction, the parameter iterative optimization of the building structure quality risk prediction model is achieved.

[0077] In the specific implementation, first determine the training cost result corresponding to the model iteration instruction of the feature space transformation and the training cost result corresponding to the model iteration instruction of the building structure quality risk prediction. For feature space transformation, if there is a lot of data distortion or the structural relationship cannot be accurately converted during the conversion from the first feature space to the second feature space or the reverse conversion process when the feature space transformation is performed on the structural data of the commercial center and the office building, the training cost result will be high. For the building structure quality risk prediction, if the prediction result deviates greatly from the actual risk situation when the risk prediction is performed on the structural data of the commercial center and the office building, the training cost result will also be high.

[0078] In the specific implementation, the training cost result corresponding to the model iteration instruction of the feature space transformation and the training cost result of the model iteration instruction of the building structure quality risk prediction are weighted and calculated. For example, according to the characteristics of the building structure data and the focus of the model, the training cost result of the feature space transformation is assigned a weight of 0.4, and the training cost result of the building structure quality risk prediction is assigned a weight of 0.6. Then, the parameters of the building structure quality risk prediction model are iteratively optimized based on the calculation results. For the weights of the neural network layer in the model, if the calculation results show that the inaccurate risk prediction is mainly due to the unreasonable weight distribution of the structural features, such as the low weight of the large-span roof structure feature of the commercial center, resulting in risk assessment deviation, then increase its weight. Repeat this process until the building structure quality risk prediction model meets the training termination conditions. For example, when the accuracy of the prediction result reaches more than 90% and the training cost result is lower than the preset threshold, it is considered that the model has been trained and can be used for the actual building structure quality risk prediction task.

[0079] In this way, by continuously adjusting the model parameters, iterative optimization of the model based on feature space transformation and building structure quality risk prediction is achieved, so that the model can maintain high prediction accuracy and robustness when processing building structure data of different types and complexities, and can ensure the effectiveness and reliability of the model in practical applications.

[0080] Thus, based on the above S201 to S204, the target prediction model is trained and obtained. Then, the target building structure data is processed using the target prediction model to output the building structure quality risk prediction result.

[0081] In addition, this application also provides another method for predicting building structure quality risks.

[0082] Specifically, another method for predicting building structure quality risk in an embodiment of the present application may include the following steps: running a model iteration instruction for feature space transformation based on sample building structure data scheduled during parameter iterative optimization of a building structure quality risk prediction model; based on the sample building structure data after feature space transformation generated by the model iteration instruction for feature space transformation, calling the building structure quality risk prediction model to run the model iteration instruction for building structure quality risk prediction, wherein the model iteration instruction for building structure quality risk prediction is used to iteratively enhance the prediction performance of the building structure quality risk prediction model for building structure quality risk structure; based on the model iteration instruction for feature space transformation and the model iteration instruction for building structure quality risk prediction, performing parameter iterative optimization on the building structure quality risk prediction model to generate a trained building structure quality risk prediction model; determining any input target building structure data, the target building structure data including at least one building structure data generated by structural extraction from a target building information model; inputting the target building structure data into the building structure quality risk prediction model to generate a building structure quality risk prediction result for the target building structure data.

[0083] The specific implementation methods of the above steps are introduced below.

[0084] After completing the feature space transformation, the sample building structure data after the transformed feature space is obtained. Taking a multi-story residential building as an example, after the conversion from the first feature space to the second feature space, the basic structure data, the main structure data and the floor structure data are all in the new feature space. At this time, the building structure quality risk prediction model is called to process these transformed sample building structure data.

[0085] For the transformed residential building structure data, the building structure quality risk prediction model first encodes it. In this process, the model will convert the data of each structural part of the residential building into a form suitable for model calculation according to its own algorithm structure. For example, the foundation bearing capacity data in the foundation structure can be encoded into a specific numerical vector, which contains the relationship information between the foundation bearing capacity and other related structural parameters. The beam and column strength data in the main structure will also be encoded into a numerical representation related to the overall stability of the building.

[0086] Based on the sample building structure characteristics after this encoding representation, the building structure quality risk training results of the sample building structure data after the feature space transformation are determined. The model will conduct risk assessment on each structural part of the residential building according to the pre-set risk assessment criteria. For example, for the foundation part, if the encoded foundation bearing capacity data is compared with the overall weight of the building and the expected load, the model will mark the corresponding risk level. For the beam and column part, if its strength coding value shows a risk after taking into account factors such as the age of the building and the extreme loads that it may bear, it will also be accurately assessed.

[0087] Then, based on the training results of the building structure quality risk of the sample building structure data after the feature space transformation, the training cost results of the building structure quality risk prediction model when running the model iteration instructions of the building structure quality risk prediction are generated. This training cost result reflects the performance of the model in the current iteration process. If in the risk assessment of residential buildings, the model accurately identifies the quality risks that may be caused by uneven foundation settlement in the foundation structure, but there is a certain deviation in the long-term durability risk assessment of the beam-column structure, then this training cost result will reflect the direction and degree of further improvement of the model in this regard.

[0088] Similarly, the building structure quality risk prediction model will perform similar operations on the sample building structure data after the transformation feature space of the commercial building. The structure of the commercial building is more complex and its risk factors are more diverse. In the encoding representation process, the model needs to take into account the interaction of more structural and non-structural factors. For example, when encoding the frame structure of a commercial building, it is necessary not only to consider the structural strength and stability of the beams and columns, but also to consider the dynamic impact of commercial activities on the building structure, such as changes in load distribution when the crowd is dense. When determining the training results of the building structure quality risk, the model needs to comprehensively evaluate the structural risks of the commercial building in different operating scenarios, such as the fire resistance of the structure in the fire scenario, the seismic performance of the structure in the earthquake scenario, etc. The training cost results generated based on these evaluation results can provide an accurate basis for further optimization of the model.

[0089] Furthermore, after obtaining the training cost results related to the model iteration instructions of feature space transformation and the model iteration instructions of building structure quality risk prediction, the parameters of the building structure quality risk prediction model are iteratively optimized.

[0090] Take the processing results of sample building structure data of multi-story residential buildings and commercial buildings as an example. If, in the processing of multi-story residential building structure data, it is found that the model has a deviation in the assessment of foundation settlement risk due to an unreasonable setting of a certain parameter in the risk prediction of the foundation structure, then the relevant parameters in the building structure quality risk prediction model are adjusted according to this feedback. For example, if the weight parameter related to the foundation soil mechanics parameter is improperly set, then adjust this weight parameter to improve the accuracy of the model's assessment of foundation settlement risk. For the case of commercial buildings, it is assumed that in the building structure quality risk prediction model, the parameters related to the structural fire resistance performance assessment did not fully consider the influence of the building's internal decoration materials in the previous iteration. In this iterative optimization process, according to the training cost results obtained from the model iteration instructions of the building structure quality risk prediction, the parameters related to the structural fire resistance performance assessment are adjusted. For example, but not limited to: adding parameter items related to the flammability of decoration materials, or adjusting the existing parameter weights related to the fire spread rate.

[0091] During the entire parameter iteration optimization process, the processing results of multiple sample building structure data are comprehensively considered. In addition to residential and commercial buildings, sample data of other types of buildings such as industrial plants can also be included. For example, there may be special equipment loads in industrial plants that affect the building structure. If the distribution of such special loads and their impact on structural risks are not well considered in the previous model, the model parameters are adjusted according to the relevant training cost results to ensure that the model can accurately assess the structural quality risks of industrial plants under the operating status of equipment.

[0092] With continuous iterative optimization, the parameters of the building structure quality risk prediction model are continuously adjusted until the model can achieve satisfactory prediction performance when processing various sample building structure data. When the prediction performance of the model reaches the set standard, for example, in the risk prediction of a large number of different types of building structure data, the accuracy reaches a high level and the error rate is controlled within an acceptable range, it is considered that a trained building structure quality risk prediction model has been generated. This trained model can accurately predict the building structure quality risk, whether it is for the building structure data corresponding to various types of buildings such as new residential buildings, commercial buildings or industrial plants.

[0093] Furthermore, the predetermined target building structure data is input into the trained building structure quality risk prediction model. Taking a new high-rise residential building as an example, when the target building structure data (including building foundation structure data, building main structure data and building floor structure data) is input into the model, the model first processes these data.

[0094] For building infrastructure data, the model evaluates whether the foundation type and size can meet the load requirements of high-rise residential buildings based on the knowledge and algorithms accumulated during the previous training process. If the design value of the foundation's bearing capacity is risky after considering the self-weight of the high-rise residential building, the load of residents, and possible wind loads and earthquake loads, the model will mark the corresponding risk level. For example, if under certain geological conditions, the pile length and pile diameter design of the pile foundation may not be able to effectively resist the horizontal load during an earthquake, the model will mark the earthquake risk of the foundation structure as a higher level.

[0095] For the main structural data of the building, the model will analyze whether the structural dimensions and material properties of the columns and beams can ensure the overall stability of the building. For example, if the diameter of some columns in a high-rise residential building is too small, there may be a risk of compressive instability considering the building height and upper load. The model will evaluate the structural stability risk of these columns and give corresponding risk warnings. At the same time, for the structure of the beam, if the height and width of the beam are not designed reasonably, resulting in insufficient bending resistance, the model will also identify this risk and conduct a quantitative assessment.

[0096] For building floor structure data, the model will check whether the floor thickness meets the requirements for residential use and the safety of the staircase structure. If the floor thickness is too thin, it may not be able to withstand the loads generated by furniture placement and human activities, and the model will mark the structural safety risk of the floor. For staircase structures, if the structural form may pose a risk of crowd congestion during emergency evacuation, the model will also give corresponding risk assessment results.

[0097] Similarly, the model will conduct a comprehensive risk assessment for the target building structure data of the commercial complex. When evaluating the large-span structure of the commercial complex, the model will judge its structural safety risk under normal use and special load conditions (such as concentrated crowd loads during large-scale events) based on the structural form and material properties. For the structure of the personnel evacuation passage, the model will comprehensively consider factors such as the width of the evacuation stairs, the length and layout of the evacuation passage, and evaluate the risk of personnel evacuation in the event of a fire or other emergency.

[0098] Therefore, the building structure quality risk prediction model is used to generate building structure quality risk prediction results for different types of target building structure data. These results can provide valuable references for the design, construction and operation of construction projects, and help relevant personnel to promptly discover and solve quality risk problems in building structures.

[0099] Among them, when processing the iterative optimization of parameters of the building structure quality risk prediction model, it involves the model iteration instructions of running feature space transformation on the sample building structure data. The following is a detailed scenario example.

[0100] For the sample building structure data scheduled in the parameter iteration optimization process of the building structure quality risk prediction model, take the data of various building types provided by a large construction company as an example. First, let's look at the operations related to the first sample building structure data. Take the building structure data of a high-rise office building as the first sample building structure data, which comes from the sample building information model data with the first feature space. The sample building information model data of this office building contains a lot of detailed information about the building structure, such as the layout, pile diameter, pile length and other data of the pile foundation in the foundation structure, the spacing, cross-sectional dimensions, and concrete strength grade of the frame columns in the main structure, and the span, cross-sectional shape, reinforcement and other data of the frame beam, as well as information about the floor thickness, floor height, and staircase structure of the floor structure. These data are in the first feature space as a whole, and this first feature space has its own feature representation. For example, the foundation structure data may exist in the form of a simple numerical combination, without establishing a complex connection with other structural or non-structural factors.

[0101] The present invention runs the model iteration instruction of the first feature space transformation from the first feature space to the second feature space for the first example building structure data of the high-rise office building. The second feature space is an ideal feature space set to enable the building structure quality risk prediction model to be better adapted when performing risk prediction. In this transformation process, for the infrastructure data, the pile foundation data originally simply represented in the first feature space will be deeply integrated with the geological survey data of the site where the office building is located when it is converted to the second feature space. For example, the pile diameter and pile length of the pile foundation will establish an association relationship with geological factors such as the bearing capacity of the stratum and the groundwater level, forming a new representation form characterized by the relationship between geology and infrastructure. For the frame column data in the main structure, in the first feature space, its cross-sectional dimensions and concrete strength grade are independently represented, and when it is converted to the second feature space, it will be combined with external load factors such as the wind load and earthquake fortification intensity of the office building. For example, the cross-sectional dimensions of the frame column will be re-quantified according to the wind load calculation results and seismic requirements, and the concrete strength grade will also be associated with the force performance of the structure under different load combinations. For the floor thickness data in the floor structure, it is just a simple value in the first feature space. When it is converted to the second feature space, it will be combined with factors such as the functional zoning and personnel load distribution inside the office building. For example, if a floor is an office area with a large and unevenly distributed personnel load, the representation of the floor thickness will be adjusted in combination with this load distribution to better reflect its impact on the quality risk of the building structure.

[0102] Next, let's look at the second example of building structure data related operations. Take the building structure data of a large commercial center as the second example building structure data, which comes from the sample building information model data with the second feature space. The building structure information model data of this commercial center covers complex structural information, such as the thickness of the raft foundation of the foundation structure, the reinforcement situation and the relationship with the underground parking structure, the special structural form of the giant columns in the main structure, the combination of steel structure and concrete structure, and the structural parameters of the large-span steel structure roof of the atrium. These data are initially in the second feature space, and the representation of this second feature space is relatively complex, containing more associations between structural and non-structural factors.

[0103] The present invention is directed to the second sample building structure data of this large commercial center, and runs the model iteration instruction of the second feature space transformation from the second feature space to the first feature space. In this process, for the raft foundation data of the foundation structure, when converting from the second feature space to the first feature space, it is necessary to simplify the representation of the complex relationship between it and the underground parking structure, etc. For example, the original representation form of the complex relationship between the thickness of the raft foundation and the column grid structure of the underground parking lot, the vehicle load transfer path, etc. in the second feature space will be simplified to the representation form that only considers the basic relationship between the thickness of the raft foundation itself, the reinforcement and the bearing capacity of the foundation when converting to the first feature space. For the giant columns in the main structure and the large-span steel structure roof structure parameters of the atrium, when converting from the second feature space to the first feature space, the association with non-structural factors such as the internal commercial layout of the commercial center and the evacuation of people will be weakened. For example, the structural form of the giant column may be associated with the atrium space layout of the commercial center, the distribution of surrounding shops, etc. in the second feature space. When converting to the first feature space, it will focus on its own structural mechanical properties, such as the representation of basic structural properties such as compression resistance and bending resistance.

[0104] The model iteration instruction of the first feature space transformation here is of great significance. It can enhance the performance of feature space transformation from the first feature space to the second feature space. For example, when converting a high-rise office building from the first feature space to the second feature space, this instruction can ensure the accuracy and effectiveness of the basic structure, main structure and floor structure data during the conversion process. It uses a series of algorithms and rules to make each structural data better match the requirements of the target feature space, thereby improving the overall feature space transformation performance.

[0105] The model iteration instructions for the second feature space transformation perform a feature space transformation from the second feature space to the first feature space, thereby synergistically enhancing the performance of the feature space transformation from the first feature space to the second feature space. Taking the sample building structure data of a large commercial center and a high-rise office building as an example, when the structural data of the commercial center is transformed from the second feature space to the first feature space, the algorithms and rules followed in this process can provide reference and supplement for the transformation of the office building from the first feature space to the second feature space. For example, in terms of processing the relationship between structural data and external loads, some processing methods of the commercial center when simplifying the association between structural and non-structural factors can provide optimization ideas for the office building when establishing the relationship between structural data and external loads, thereby synergistically enhancing the performance of the feature space transformation from the first feature space to the second feature space.

[0106] Finally, the sample building structure data after the feature space transformation includes the sample building structure data after the feature space transformation generated by the model iteration instruction of running the first feature space transformation from the first feature space to the second feature space for the first sample building structure data. For example, after such transformation, the basic structure, main structure and floor structure data of a high-rise office building are all in the new feature space, and the representation form and association relationship of these data have changed, forming a new form of building structure data. Such transformed sample building structure data will provide an important data basis for the operation of the subsequent building structure quality risk prediction model.

[0107] For the first example building structure data (such as the building structure data of the high-rise office building mentioned earlier) scheduled during the parameter iterative optimization process based on the building structure quality risk prediction model, the process of running the model iteration instruction of the first feature space transformation from the first feature space to the second feature space is as follows.

[0108] First, the first example building structure data is transformed from the first feature space to the second feature space. For high-rise office buildings, the basic structure data (such as the layout, pile diameter, pile length, etc. of the pile foundation) exists in a relatively simple and independent form in the first feature space, and the correlation with other factors is weak. When transforming to the second feature space, it is necessary to combine with more factors. For example, the pile diameter and pile length of the pile foundation should not only consider their own values, but also be associated with the geological conditions of the site where the office building is located (such as the lithology of the stratum, groundwater level, soil bearing capacity, etc.). The data such as the frame column spacing, cross-sectional dimensions and concrete strength grade in the main structure are independently represented in the first feature space. When transforming to the second feature space, they should be combined with factors such as the wind load, seismic load and overall building height faced by the office building. The data such as the floor thickness, floor height and staircase structure form of the floor structure should be linked to the functional layout inside the office building (such as the distribution of office areas, equipment rooms, etc.) and the distribution of personnel loads.

[0109] Based on the first sample building structure data after the feature space is transformed, the first feature space discrimination iterative sub-instruction corresponding to the model iteration instruction of the first feature space transformation is run to generate the training cost result corresponding to the first feature space discrimination iterative sub-instruction. In this process, the first feature space discrimination iterative sub-instruction aims to iteratively strengthen the discrimination performance of the building structure data generated by the feature space transformation from the first feature space to the second feature space. For the transformed foundation structure data of the high-rise office building, the discrimination iterative sub-instruction will check the rationality of the pile foundation after the association with the geological conditions. For example, if the pile diameter and pile length of the pile foundation are adjusted according to the geological conditions, whether they can meet the bearing requirements of the office building under different working conditions (such as normal use, earthquake action, etc.). If under certain geological conditions, the pile length is too short or the pile diameter is too small, resulting in insufficient bearing capacity, the discrimination iterative sub-instruction should be able to identify this situation. For the main structure, it is judged whether the structural performance of the frame column after being associated with factors such as wind load, seismic load and building height is reasonable. If the cross-sectional size of the frame column does not conform to the principles of structural mechanics after considering these factors, the discrimination iterative sub-instruction should be able to detect it. For floor structures, determine the safety of the floor thickness after being associated with the functional layout and the distribution of personnel loads. If the floor thickness does not meet the requirements after considering the load of the concentrated personnel area, the judgment iterative sub-instruction should be able to determine. Based on these judgment results, the training cost results are generated. If the judgment performance is good, the training cost result is low; if there are many judgment errors, the training cost result is high.

[0110] Next, the first example building structure data after the feature space transformation is transformed from the second feature space to the first feature space. For the infrastructure data of high-rise office buildings, the state deeply associated with geological conditions is restored to a relatively independent representation form. For example, the pile foundation representation after considering factors such as stratum lithology and groundwater level is restored to a form that only focuses on its own basic parameters. In the main structure, the frame column data associated with wind load, seismic load and building height is restored to the original relatively independent size and strength grade representation. In terms of floor structure, data such as floor thickness associated with functional layout and personnel load distribution are restored to the initial simple representation.

[0111] The first feature space transformation iterative learning sub-instruction corresponding to the model iteration instruction of the first feature space transformation is run based on the first sample building structure data of the feature space inverse transformation and the first sample building structure data before the feature space transformation, and the training cost result of the first feature space transformation iterative learning sub-instruction is generated. The first feature space transformation iterative learning sub-instruction is used to enhance the validity of the building structure data obtained by the feature space transformation from the first feature space to the second feature space so as to improve the probability of false identification in the first feature space discrimination iterative sub-instruction. For high-rise office buildings, the inverse transformed infrastructure data is compared with the original infrastructure data. If some adjustments to the pile foundation data are found to have a large deviation from the original data after the inverse transformation during the feature space transformation process, and this deviation may lead to structural safety risks, then the transformation iterative learning sub-instruction needs to adjust the relevant transformation algorithm. For the main structure, if the inverse transformed frame column data is compared with the original data, it is found that the previous transformation causes a large error in the evaluation of the structural performance, and the transformation iterative learning sub-instruction needs to be optimized. For the floor structure, if the data such as the floor thickness have unreasonable differences from the original data after the inverse transformation, which affects the judgment of the structural safety, the transformation algorithm also needs to be improved. The training cost result is generated based on the results of these comparisons and adjustments. If a large number of adjustments to the transformation algorithm are required, the training cost result is high; if only a small amount of optimization is required, the training cost result is low. This training cost result will provide an important basis for subsequent model optimization, so as to continuously improve the accuracy and effectiveness of the feature space transformation from the first feature space to the second feature space, thereby improving the performance of the building structure quality risk prediction model.

[0112] For the second sample building structure data (such as the building structure data of the large commercial center mentioned earlier) scheduled during the parameter iterative optimization process based on the building structure quality risk prediction model, the model iteration instruction of running the second feature space transformation from the second feature space to the second feature space is as follows. The detailed process is as follows.

[0113] First, the second sample building structure data is transformed from the second feature space to the first feature space. For large commercial centers, the data such as the thickness of the raft foundation, the reinforcement, and the correlation with the underground parking structure in the foundation structure have a complex representation in the second feature space, including the correlation with many structural and non-structural factors. When transforming to the first feature space, these data need to be simplified. For example, the representation of the complex relationship between the thickness of the raft foundation and the column grid structure of the underground parking lot and the vehicle load transfer path in the second feature space should be converted into a representation that only considers the basic relationship between the thickness of the raft foundation itself, the reinforcement and the bearing capacity of the foundation. The special structural form of the giant column in the main structure, the combination of steel structure and concrete structure, etc., are related to non-structural factors such as the internal commercial layout and crowd evacuation of the commercial center in the second feature space. When transforming to the first feature space, it is necessary to focus on its own structural mechanical properties, such as the representation of the basic structural properties such as the compression and bending capacity of the giant column, and weaken the correlation with non-structural factors. The structural parameters of the large-span steel structure roof of the atrium are related to the overall spatial layout, lighting design and other factors of the commercial center in the second characteristic space. When converted to the first characteristic space, they should be simplified to focus only on its own structural characteristics, such as the structural stability of the roof, the internal force of the rods and other basic parameters.

[0114] Based on the second sample building structure data after the feature space is transformed, the second feature space discrimination iterative sub-instruction corresponding to the model iteration instruction of the second feature space transformation is run to generate the training cost result corresponding to the second feature space discrimination iterative sub-instruction. In this process, the second feature space discrimination iterative sub-instruction is used to iteratively enhance the discrimination performance of the building structure data obtained by the feature space transformation from the second feature space to the first feature space. For the transformed foundation structure data of the large commercial center, the discrimination iterative sub-instruction will check the rationality of the simplified representation of the raft foundation. For example, when only considering the basic relationship between the thickness, reinforcement and foundation bearing capacity of the raft foundation itself, it is necessary to judge whether this simplified representation can accurately reflect the performance of the raft foundation under different working conditions (such as normal use, uneven settlement, etc.). If the reinforcement of the raft foundation cannot meet the foundation bearing capacity requirements after the simplified calculation, the discrimination iterative sub-instruction should be able to identify this situation. For the main structure, it is judged whether the structural performance of the giant column is reasonable after paying attention to its own structural mechanical properties. If the calculation of the compressive or bending capacity of the giant column does not conform to the principles of structural mechanics after the simplified association, the discrimination iterative sub-instruction should be able to detect it. For the large-span steel structure roof of the atrium, determine whether its structural stability meets the requirements after being simplified to its own structural characteristics. If the internal forces of the roof members lead to insufficient structural stability after simplified calculation, the judgment iterative sub-instruction should be able to determine it. Based on these judgment results, the training cost results are generated. If the judgment performance is good, the training cost result is low; if there are many judgment errors, the training cost result is high.

[0115] Next, the second sample building structure data after the feature space transformation is transformed from the first feature space to the second feature space. For the basic structural data of large commercial centers, the representation form that only considers its own basic relationships is restored to a representation form that has complex relationships with underground parking column grid structures, vehicle load transfer paths, etc. In the main structure, the giant column data that focuses on its own structural mechanical properties is restored to a representation form that is associated with non-structural factors such as the internal commercial layout and crowd evacuation of the commercial center. The structural parameters of the atrium's large-span steel structure roof are restored from only focusing on its own structural characteristics to a representation form that is associated with the overall spatial layout of the commercial center, lighting design and other factors.

[0116] The second feature space transformation iterative learning sub-instruction corresponding to the model iteration instruction of the second feature space transformation is run based on the second sample building structure data of the feature space inverse transformation and the second sample building structure data before the feature space transformation, and the training cost result of the second feature space transformation iterative learning sub-instruction is generated. The second feature space transformation iterative learning sub-instruction is used to enhance the validity of the building structure data obtained by the feature space transformation from the second feature space to the first feature space so as to improve the probability of false identification in the second feature space discrimination iterative sub-instruction. For large commercial centers, the inverse transformed basic structure data is compared with the original basic structure data. If some adjustments to the raft foundation data are found to have a large deviation from the original data after the inverse transformation during the feature space transformation process, and this deviation may cause structural safety risks, then the transformation iterative learning sub-instruction needs to adjust the relevant transformation algorithm. For the main structure, if the inverse transformed giant column data is compared with the original data, it is found that the previous transformation causes a large error in the evaluation of the structural performance, and the transformation iterative learning sub-instruction needs to be optimized. For the large-span steel structure roof of the atrium, if its structural parameters have unreasonable differences from the original data after the inverse transformation, which affects the judgment of the structural safety, the transformation algorithm also needs to be improved. The training cost result is generated according to the results of these comparisons and adjustments. If a large amount of adjustment of the transformation algorithm is required, the training cost result is high; if only a small amount of optimization is required, the training cost result is low. This training cost result will provide an important basis for subsequent model optimization, so as to continuously improve the accuracy and effectiveness of the feature space transformation from the second feature space to the first feature space, thereby improving the performance of the building structure quality risk prediction model. Through such operations, the present invention can effectively perform feature space transformation processing on building structure data such as large commercial centers, laying the foundation for the accurate operation of the building structure quality risk prediction model. In actual operation, the present invention needs to accurately calculate and distinguish each data point and each structural element to ensure that the processing of the entire building structure data meets the requirements of the building structure quality risk prediction, and can continuously optimize the performance of the model so that it can accurately predict the building structure quality risk when processing different types of building structure data.

[0117] Therefore, by introducing the model iteration instructions of feature space transformation, the sample building structure data is effectively transformed in feature space, so that the building structure quality risk prediction model can better adapt to building structure data of different types and complexities, thereby improving the accuracy and robustness of the prediction.

[0118] In some optional embodiments, based on the sample building structure data after the transformed feature space generated by the model iteration instruction of the feature space transformation, the building structure quality risk prediction model is called to run the model iteration instruction of the building structure quality risk prediction, including: using the building structure quality risk prediction model to encode and represent the multiple sample building structure data after the transformed feature space generated by the model iteration instruction of the feature space transformation; wherein the multiple sample building structure data after the transformed feature space is the building structure data of the second feature space; based on the sample building structure features after the encoded representation, determining the building structure quality risk training results of the multiple sample building structure data after the transformed feature space; based on the building structure quality risk training results of the multiple sample building structure data after the transformed feature space, generating the training cost result when the building structure quality risk prediction model runs the model iteration instruction of the building structure quality risk prediction.

[0119] In some optional embodiments, the building structure quality risk prediction model is used to encode and represent the sample building structure data after multiple transformed feature spaces generated by the model iteration instructions of the feature space transformation, including: inputting the sample building structure data after multiple transformed feature spaces generated by the model iteration instructions of the feature space transformation into the encoding module of the building structure quality risk prediction model; using the encoding module to encode and represent the sample building structure data after multiple transformed feature spaces; using the encoding module to cross-fuse the building structure features corresponding to the sample building structure data after multiple transformed feature spaces in the spatial dimension.

[0120] In summary, the advantages of the present invention are mainly reflected in the following aspects:

[0121] First, by introducing the model iteration instruction of feature space transformation, the present invention can dynamically adjust the feature space so that the building structure quality risk prediction model can better adapt to data of different types and complexity. This advantage solves the problem that the model in the prior art is difficult to handle complex and diverse data structures, and improves the accuracy and robustness of the prediction. In practical applications, this means that the present invention can more accurately assess the quality risks of various building structures and provide more reliable protection for the safety and durability of construction projects.

[0122] Secondly, the feature encoding and fusion mechanism of the present invention can capture the complex relationship between structural features, further enhancing the model's ability to understand complex structural features. This advantage enables the model to better understand the physical meaning and structural relationship behind the data when processing complex building structure data, thereby improving the accuracy and reliability of the prediction. In practical applications, this means that the present invention can more accurately predict the quality risks of complex building structures and provide a more scientific basis for the design and construction of construction projects.

[0123] Finally, by proposing a model iterative optimization method based on feature space transformation and building structure quality risk prediction, the present invention can ensure that the model maintains high prediction accuracy and robustness when processing data structures of different types and complexities. This advantage makes the model more adaptable and reliable in practical applications, and can cope with various complex and changeable situations. In practical applications, this means that the present invention can provide a more stable and reliable solution for quality risk prediction of construction projects, which helps to improve the safety level and engineering quality of the entire construction industry.

[0124] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0125] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a building structure quality risk prediction device 300.

[0126] like Figure 3 As shown, the building structure quality risk prediction device 300 may include:

[0127] The extraction module 301 is used to perform structural extraction on the pre-built target building information model to obtain a number of target building structure data, wherein the target building structure data at least includes building foundation structure data, building main structure data, and building floor structure data;

[0128] The input module 302 is used to input the target building structure data into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result. The target prediction model is trained based on multiple first building structure data samples and second building structure data samples after the feature space of the first building structure data samples is transformed.

[0129] In some embodiments, the building structure quality risk prediction device 300 further includes a training module ( Figure 3 Specifically, the training module includes the following units:

[0130] An acquisition unit, used for acquiring a plurality of first building structure data samples;

[0131] A transformation unit, used for performing feature space transformation on the first building structure data sample using a transformation iterative learning sub-instruction to obtain a second building structure data sample after the feature space transformation;

[0132] A generating unit, configured to generate, according to the first building structure data sample and the second building structure data sample, a training cost result of a model iteration instruction for running building structure quality risk prediction corresponding to a preset model;

[0133] The adjustment unit is used to adjust the weight parameters of the preset model according to the training cost result, and train the adjusted preset model until the preset training stop condition is met to obtain the trained target prediction model.

[0134] In some optional embodiments, the generation unit is specifically used to: perform an inverse transformation of the feature space of the second building structure data sample after the feature space is transformed to obtain a third building structure data sample; determine the deviation between the first building structure data sample and the third building structure data sample; when the deviation is greater than or equal to a preset threshold, determine the adjustment information of the transformation algorithm corresponding to the transformation iterative learning sub-instruction according to the deviation; optimize the transformation iterative learning sub-instruction according to the adjustment information; and generate a training cost result according to the deviation and the adjustment information.

[0135] Optionally, the first building structure data sample includes at least structural mechanics parameters.

[0136] In some optional embodiments, the transformation unit is specifically used to: when the initial state of the first building structure data sample is the first feature space, establish the association relationship between the structural mechanics parameters and the pre-set non-structural factor parameters to obtain a second building structure data sample after the feature space is transformed, the state of the second building structure data sample is the second feature space, and the second building structure data sample includes structural mechanics parameters, non-structural factor parameters and characteristics of the association relationship between the building structure and the non-structural factors; when the initial state of the first building structure data sample is the second feature space, simplify the association relationship between the building structure and the non-structural factors to obtain the second building structure data sample after the feature space is transformed, and the state of the second building structure data sample is the first feature space.

[0137] In some optional embodiments, the training module also includes an encoding unit, which is used to input multiple second building structure data samples into the encoding module of the preset model; use the encoding module to encode the multiple second building structure data samples respectively to obtain multiple sample feature vectors; use the encoding module to cross-fuse the multiple sample feature vectors in the spatial dimension.

[0138] In some optional embodiments, the adjustment unit is specifically used to: use multiple sample feature vectors as input of a preset model, train the adjusted preset model until a preset training stop condition is met, and obtain a trained target prediction model.

[0139] It should be noted that, for the convenience of description, the above device is described in various modules according to their functions. Of course, when implementing this application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0140] The device of the above embodiment is used to implement the corresponding building structure quality risk prediction method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0141] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides an electronic device.

[0142] Figure 4 A more specific schematic diagram of the hardware structure of an electronic device provided by this embodiment is shown.

[0143] The electronic device 400 may include a processor 401 and a memory 402 storing computer program instructions.

[0144] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0145] The memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 402 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 402 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 402 is a non-volatile solid-state memory.

[0146] In certain embodiments, the memory may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage media device, an optical storage media device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, typically, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present application.

[0147] The processor 401 implements any one of the building structure quality risk prediction methods in the above embodiments by reading and executing computer program instructions stored in the memory 402 .

[0148] In some examples, the electronic device 400 may further include a communication interface 403 and a bus 410. Figure 4 As shown, the processor 401, the memory 402, and the communication interface 403 are connected via a bus 410 and communicate with each other.

[0149] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0150] Bus 410 includes hardware, software or both, and couples the components of online data traffic billing equipment to each other. For example, but not limitation, bus 410 may include accelerated graphics port (AGP) or other graphics bus, enhanced industry standard architecture (EISA) bus, front-side bus (FSB), hypertransport (HT) interconnection, industry standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnect (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. Where appropriate, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0151] Exemplarily, the electronic device 400 may be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA).

[0152] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides a non-transitory computer-readable storage medium. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any of the building structure quality risk prediction methods in the above-mentioned embodiments is implemented. Examples of computer-readable storage media include non-transitory computer-readable storage media, such as portable disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, etc.

[0153] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer so that the computer and / or the processor execute the building structure quality risk prediction method. Corresponding to the execution subject corresponding to each step in each embodiment of the building structure quality risk prediction method, the processor that executes the corresponding step can belong to the corresponding execution subject.

[0154] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0155] The functional blocks shown in the structural block diagram described above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier. "Machine-readable medium" may include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memories, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0156] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0157] The above describes various aspects of the present application with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0158] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A method for predicting building structure quality risk, characterized in that: include: Performing structural extraction on a pre-constructed target building information model to obtain a plurality of target building structure data, wherein the target building structure data at least includes building foundation structure data, building main structure data, and building floor structure data; The target building structure data is input into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result, wherein the target prediction model is trained based on multiple first building structure data samples and second building structure data samples after transforming the feature space of the first building structure data samples.

2. The method according to claim 1, characterized in that Before inputting the target building structure data into the pre-trained target prediction model, the method further includes: Acquire a plurality of first building structure data samples; Using the transformation iterative learning sub-instruction to perform feature space transformation on the first building structure data sample, to obtain a second building structure data sample after the feature space transformation; Generate, based on the first building structure data sample and the second building structure data sample, a training cost result when running a model iteration instruction for building structure quality risk prediction corresponding to a preset model; The weight parameters of the preset model are adjusted according to the training cost result, and the adjusted preset model is trained until the preset training stop condition is met, so as to obtain the trained target prediction model.

3. The method according to claim 2, characterized in that The generating, according to the first building structure data sample and the second building structure data sample, a training cost result of a model iteration instruction for running building structure quality risk prediction corresponding to a preset model comprises: Performing an inverse feature space transformation on the second building structure data sample after the feature space transformation to obtain a third building structure data sample; determining a deviation between the first building structure data sample and the third building structure data sample; In the case where the deviation is greater than or equal to a preset threshold, determining adjustment information of the transformation algorithm corresponding to the transformation iterative learning sub-instruction according to the deviation; Optimizing the transformation iterative learning sub-instruction by using the adjustment information; A training cost result is generated according to the deviation and the adjustment information.

4. The method according to claim 2, characterized in that: The first building structure data sample at least includes structural mechanics parameters; The step of performing feature space conversion on the first building structure data sample to obtain a second building structure data sample after the feature space is converted includes: In the case where the initial state of the first building structure data sample is the first feature space, an association relationship between the structural mechanics parameters and the pre-set non-structural factor parameters is established to obtain a second building structure data sample after the feature space is transformed, the state of the second building structure data sample is the second feature space, and the second building structure data sample includes the structural mechanics parameters, the non-structural factor parameters, and the characteristics of the association relationship between the building structure and the non-structural factor; When the initial state of the first building structure data sample is the second feature space, the association between the building structure and the non-structural factors is simplified to obtain the second building structure data sample after the feature space is transformed, and the state of the second building structure data sample is the first feature space.

5. The method according to claim 2, characterized in that: The method further comprises: Inputting a plurality of second building structure data samples into the encoding module of the preset model; Using the encoding module to encode a plurality of second building structure data samples respectively to obtain a plurality of sample feature vectors; The encoding module is used to cross-fuse multiple sample feature vectors in the spatial dimension.

6. The method according to claim 5, characterized in that The step of adjusting the weight parameters of the preset model according to the training cost result, and training the adjusted preset model until a preset training stop condition is met to obtain a trained target prediction model includes: Taking multiple sample feature vectors as input of the preset model, the adjusted preset model is trained until the preset training stop condition is met, thereby obtaining the trained target prediction model.

7. A building structure quality risk prediction device, characterized in that: The device comprises: An extraction module is used to perform structural extraction on a pre-built target building information model to obtain a number of target building structure data, wherein the target building structure data at least includes building foundation structure data, building main structure data, and building floor structure data; An input module is used to input the target building structure data into a pre-trained target prediction model to obtain a corresponding building structure quality risk prediction result. The target prediction model is trained based on multiple first building structure data samples and second building structure data samples after transforming the feature space of the first building structure data samples.

8. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; when the processor calls the computer program instructions, the method for predicting building structure quality risks as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are called by the processor, the method for predicting building structure quality risks according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the building structure quality risk prediction method as described in any one of claims 1-6.

Images