A building information management system based on BIM technology

Abstract

The present invention relates to the technical field of digital management of building data, and particularly to a building information management system based on BIM technology. It obtains the shape modeling data of the target entity through a data parsing module, then establishes the target entity model in a three-dimensional grid space through a model building module, and then uses a data encoding module to perform encoding to obtain a three-dimensional input matrix. After that, a type recognition module uses a preset three-dimensional convolutional neural network to identify the type of the target entity. Finally, an information supplement module supplements the building information of the target entity. Compared with the prior art, the present invention completes encoding by using the coverage relationship between the three-dimensional grid space and the target entity model, so that the type of the entity can be identified through the three-dimensional convolutional neural network, and then other missing information of the target entity can be automatically supplemented based on the type of the target entity, solving the problem that the BIM technology in the prior art cannot be normally applied in the case of data loss.

Description

Technical Field

[0001] The present invention relates to the technical field of digital management of building data, and particularly to a building information management system based on BIM technology. Background Art

[0002] Building Information Modeling (BIM) is a method for comprehensively managing building projects through digital means. It is not only a design tool but also an information integration platform for the entire building life cycle. Through BIM, building professionals can share, manage, and analyze building-related data at all stages of a project, thereby improving design quality, optimizing construction processes, and reducing operating costs.

[0003] Although BIM contains rich data resources, in actual applications, the technical levels of practitioners vary, and the actual situations of building projects are complex and changeable. This has led to the inability to guarantee the integrity of information entry, and data loss has become a common problem. The lack of complete data will seriously hinder the normal application of BIM technology and limit the speed and scope of its promotion in the construction industry.

[0004] Therefore, there is an urgent need for a building information management system based on BIM technology that can work in the case of data loss. Summary of the Invention

[0005] Therefore, the present invention provides a building information management system based on BIM technology to solve the problem that BIM technology in the prior art cannot be normally applied in the case of data loss.

[0006] The present invention provides a building information management system based on BIM technology, including:

[0007] A data parsing module for obtaining shape modeling data of a target entity based on a BIM model;

[0008] A model building module for building a three-dimensional grid space and modeling in the three-dimensional grid space based on the shape modeling data to obtain a target entity model in the three-dimensional grid space;

[0009] A data encoding module for encoding according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix, where each element in the three-dimensional input matrix corresponds to a grid in the three-dimensional grid space, and the value of each element in the three-dimensional input matrix is used to represent the partial target entity model in its corresponding grid;

[0010] A type recognition module for inputting the three-dimensional input matrix into a preset three-dimensional convolutional neural network to obtain the type of the target entity output by the preset three-dimensional convolutional neural network;

[0011] An information supplement module for supplementing the building information of a target entity in a BIM system based on the type of the target entity.

[0012] The present invention also provides a preferred solution: based on a BIM model, obtaining shape modeling data of a target entity, including:

[0013] Obtaining input range selection data representing the target entity, and based on the range selection data, obtaining the BIM shape data of the target entity according to the BIM model;

[0014] Based on the BIM shape data, obtaining general model data in STEP format for the target entity;

[0015] Obtaining the entity number, the position information and shape information corresponding to each entity number in the general model data as the shape modeling data of the target entity.

[0016] The present invention also provides a preferred solution: establishing a three-dimensional grid space and performing modeling in the three-dimensional grid space based on the shape modeling data to obtain a target entity model in the three-dimensional grid space, including:

[0017] Establishing a three-dimensional grid space and obtaining the size data of the three-dimensional grid space;

[0018] Based on the size data of the three-dimensional grid space, the position information and shape information in the shape modeling data, performing coordinate transformation on the position information to obtain standard position data;

[0019] Based on the size data of the three-dimensional grid space, the position information and shape information in the shape modeling data, performing scaling on the shape information to obtain standard shape data;

[0020] Based on the entity number, the standard position information and standard shape information corresponding to each entity number, performing modeling in the three-dimensional grid space to obtain a target entity model in the three-dimensional grid space.

[0021] The present invention also provides a preferred solution: encoding according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix, including:

[0022] Obtaining a target grid, where the target grid is a grid in the three-dimensional grid space;

[0023] Calculating the spatial occupancy ratio of the target entity model in the target grid and normalizing the spatial occupancy ratio to obtain the element value of the three-dimensional input matrix corresponding to the target grid.

[0024] The present invention also provides a preferred solution: the preset three-dimensional convolutional neural network includes an input layer, at least one three-dimensional convolutional layer, at least one pooling layer, at least one fully connected layer, and an output layer. The input layer is located at the first layer of the preset three-dimensional convolutional neural network. A three-dimensional convolutional layer is connected after the input layer. The output layer is located at the last layer of the preset three-dimensional convolutional neural network. A fully connected layer is connected before the output layer. The other three-dimensional convolutional layers, pooling layers, and fully connected layers are sequentially connected between the three-dimensional convolutional layer after the input layer and the fully connected layer before the output layer in an arranged and combined manner.

[0025] The present invention also provides a preferred solution: the three-dimensional convolutional layer after the input layer includes a shape three-dimensional convolutional layer, a long-edge convolutional layer, a wide-edge convolutional layer, a high-edge convolutional layer, and a fusion layer. The input ends of the shape three-dimensional convolutional layer, the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer are all connected to the input layer. The output ends of the shape three-dimensional convolutional layer, the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer are all connected to the input end of the fusion layer. Among them, the shape three-dimensional convolutional layer is used to detect the shape features of the target entity model. The long-edge convolutional layer is used to detect the shape edge features of the target entity model in the long dimension. The wide-edge convolutional layer is used to detect the shape edge features of the target entity model in the wide dimension. The high-edge convolutional layer is used to detect the shape edge features of the target entity model in the high dimension. The fusion layer is used to respectively superimpose and fuse the output results of the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer into the output result of the shape three-dimensional convolutional layer in the three dimensions of length, width, and height, and output the first convolutional three-dimensional matrix.

[0026] The present invention also provides a preferred solution: the sizes of the three-dimensional grid space in the three dimensions of length, width, and height are all n. The sizes of the convolutional kernels of the shape three-dimensional convolutional layer in the three dimensions of length, width, and height are all a, where both a and n are positive integers and a is less than n. The convolutional kernel of the shape three-dimensional convolutional layer is used to move step by step in the three dimensions of length, width, and height of the input three-dimensional matrix. The shape three-dimensional convolutional layer is used to output the shape feature three-dimensional matrix. The sizes of the convolutional kernels of the long-edge convolutional layer in the three dimensions of length, width, and height are n, a, and a respectively. The convolutional kernel of the long-edge convolutional layer is used to move step by step in the two dimensions of width and height of the input three-dimensional matrix. The long-edge convolutional layer is used to output the long-edge feature two-dimensional matrix. The sizes of the convolutional kernels of the wide-edge convolutional layer in the three dimensions of length, width, and height are a, n, and a respectively. The convolutional kernel of the wide-edge convolutional layer is used to move step by step in the two dimensions of length and height of the input three-dimensional matrix. The long-edge convolutional layer is used to output the wide-edge feature two-dimensional matrix. The sizes of the convolutional kernels of the high-edge convolutional layer in the three dimensions of length, width, and height are a, a, and n respectively. The convolutional kernel of the high-edge convolutional layer is used to move step by step in the two dimensions of length and width of the input three-dimensional matrix. The high-edge convolutional layer is used to output the high-edge feature two-dimensional matrix.

[0027] The present invention also provides a preferred solution: The fusion layer is used to stack and fuse the long-edge feature two-dimensional matrix, the wide-edge feature two-dimensional matrix, and the high-edge feature two-dimensional matrix into the shape feature three-dimensional matrix in the three dimensions of length, width, and height respectively, and output the first convolutional three-dimensional matrix. The corresponding formula is:

[0028]

[0029] Wherein, represents the value of the element at coordinates i, j, and k in the three dimensions of length, width, and height in the first convolutional three-dimensional matrix, represents the value of the element at coordinates i, j, and k in the three dimensions of length, width, and height in the shape feature three-dimensional matrix, represents the value of the element at coordinates j and k in the two dimensions of width and height in the long-edge feature two-dimensional matrix, represents the value of the element at coordinates i and k in the two dimensions of length and height in the wide-edge feature two-dimensional matrix, represents the value of the element at coordinates i and j in the two dimensions of length and width in the high-edge feature two-dimensional matrix. w0 represents the weight value between the node corresponding to the element in the fusion layer and the node corresponding to the element in the shape three-dimensional convolutional layer. w1 represents the weight value between the node corresponding to the element in the fusion layer and the node corresponding to the element in the long-edge convolutional layer. w2 represents the weight value between the node corresponding to the element in the fusion layer and the node corresponding to the element in the wide-edge convolutional layer. w3 represents the weight value between the node corresponding to the element in the fusion layer and the node corresponding to the element in the high-edge convolutional layer. b i,j,k represents the bias value of the node corresponding to the element in the fusion layer. f() is a preset activation function, and m is the number of elements in the shape feature three-dimensional matrix in the length dimension.

[0030] The present invention also provides a preferred solution: The output layer includes a plurality of output nodes. The plurality of output nodes are respectively connected to a plurality of nodes in the fully connected layer adjacent to the output layer. The number of output nodes is equal to the preset number of types of target entities. Each output node corresponds to a preset type, and the output node is used to output the probability value that the type of the target entity is its corresponding preset type.

[0031] The present invention also provides a preferred solution: The building information includes material information, attribute information, cost information, spatial information, and structural information; Based on the type of the target entity, supplement the building information of the target entity in the BIM system, including:

[0032] Retrieve entities of the same type as the target entity type in the BIM system;

[0033] Obtain the material information of the entities of the same type as the target entity type in the BIM system as the material information of the target entity;

[0034] Based on the material information of the target entity, supplement the attribute information, cost information, spatial information, and structural information of the target entity.

[0035] The beneficial effects of adopting the above embodiments are as follows:

[0036] The present invention provides a building information management system based on BIM technology. Through a data parsing module, shape modeling data of a target entity is obtained based on a BIM model. Then, a three-dimensional grid space is established through a model building module and modeling is performed in the three-dimensional grid space to obtain a target entity model. Next, a data encoding module encodes according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix. After that, a type recognition module uses the three-dimensional input matrix and a preset three-dimensional convolutional neural network to identify the target entity type. Finally, an information supplement module supplements the building information of the target entity in the BIM system based on the target entity type. Compared with the prior art, the present invention is based on the most important and basic shape information in the BIM system, that is, shape modeling data, performs modeling in a three-dimensional grid space, and uses the coverage relationship between the three-dimensional grid space and the target entity model to complete the encoding of the entity shape. In this way, the entity type can be identified through a three-dimensional convolutional neural network, and then other missing information of the target entity can be automatically supplemented based on the type of the target entity. In this way, the entire BIM system can temporarily start working based on the automatically filled building information and make changes when subsequent information is entered, solving the problem that the BIM technology in the prior art cannot be normally applied in the case of data loss. Description of the Drawings

[0037] Figure 1 It is a system architecture diagram of an embodiment of the building information management system based on BIM technology provided by the present invention;

[0038] Figure 2 It is a structural schematic diagram of a simplified first three-dimensional convolutional layer in the present invention. Detailed Embodiments

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0040] Combined Figure 1 As shown, a specific embodiment of the present invention discloses a building information management system based on BIM technology, including:

[0041] A data parsing module 110, configured to obtain shape modeling data of a target entity based on a BIM model;

[0042] A model building module 120, configured to build a three-dimensional grid space and model in the three-dimensional grid space based on the shape modeling data to obtain a target entity model in the three-dimensional grid space;

[0043] A data encoding module 130, configured to perform encoding according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix. Each element in the three-dimensional input matrix corresponds to a grid in the three-dimensional grid space, and the value of each element in the three-dimensional input matrix is used to represent the partial target entity model within its corresponding grid;

[0044] A type recognition module 140, configured to input the three-dimensional input matrix into a preset three-dimensional convolutional neural network to obtain the type of the target entity output by the preset three-dimensional convolutional neural network;

[0045] An information supplement module 150, configured to supplement the building information of the target entity in the BIM system based on the type of the target entity.

[0046] In the above process, the three-dimensional grid space is a virtual space established in a computer system and divided into multiple grids. By building a target entity model in the three-dimensional grid space, the shape of the target entity can be represented in the form of a vector, that is, the three-dimensional input matrix.

[0047] The three-dimensional input matrix is a combination of numbers in three dimensions, which can be represented by a two-dimensional matrix composed of multiple one-dimensional vectors. For example:

[0048]

[0049] The above formula can be regarded as a 2×2×2 three-dimensional input matrix, corresponding to a 2×2×2 three-dimensional grid space. Each element in the above three-dimensional input matrix corresponds to a grid in a three-dimensional grid space, where 0 indicates that there is no entity in the grid, and 1 indicates that there is an entity in the grid. Therefore, the above formula can also be understood as that there is a 1×1×1 block model in the 2×2×2 three-dimensional grid space.

[0050] Through the above means, the shape information of a target entity can be represented in digital form and can then be used as the input of a neural network. Utilizing the ability of the neural network to handle complex non-linear problems, the type of the target entity can be recognized based on the shape information, such as pipelines, walls, or steel bar support structures, etc.

[0051] The shape information is the basic carrier of other data information in the BIM model. In most cases, the shape information will not be missing in the BIM model. After identifying the type of the target entity through the shape information, other information can be supplemented based on the type using other existing relevant materials. For example, if it is known that the target entity is an external wall pipeline, its material is most likely metal or plastic. Based on this inference, building information such as the cost and stress of the pipeline can be further automatically filled to ensure the safety and reliability of the pipeline during the design and construction stages. This process of type inference and selection based on shape information can help optimize the design and construction processes of building projects and improve the overall quality and efficiency.

[0052] Furthermore, in a preferred embodiment, the steps performed by the above data parsing module 110 are as follows: Based on the BIM model, obtain the shape modeling data of the target entity, specifically including:

[0053] Obtain the input range selection data representing the target entity. Based on the range selection data, obtain the BIM shape data of the target entity according to the BIM model;

[0054] Based on the BIM shape data, obtain the general model data in STEP format of the target entity;

[0055] Obtain the entity number, the position information and shape information corresponding to each entity number in the general model data as the shape modeling data of the target entity.

[0056] The range selection data in the above process is the range of the entity for which data supplementation is desired, specified manually in the BIM model.

[0057] The STEP (Standard for the Exchange of Product model data) format in the above process is part of the ISO 10303 standard and is used to represent the shape of product data. It is a general data format widely used in the fields of CAD (Computer Aided Design) and BIM (Building Information Modeling). STEP files mainly use text format to facilitate the description of complex three-dimensional models and related information.

[0058] STEP files usually include the following main parts:

[0059] 1. Header Section: Contains descriptive information about the file, such as file name, creation time, author, etc. And specifies the model standard used by the file (e.g., IFC4).

[0060] 2. Data Section: Contains the actual model data, represented in the form of STEP entities. Each entity has a unique number (#1, #2, etc., i.e., entity number).

[0061] 3. Entities and Attributes: Specific information for each entity, such as type, attribute values, relationships, etc. Entities can contain geometric representations, text attributes, relationship links, etc. information.

[0062] To extract specific information (such as entity numbers, location information, and shape information) from a STEP file, the following steps can be taken:

[0063] 1. Identify entity numbers:

[0064] Entities in the STEP file are identified by unique numbers (e.g., #1, #2). These numbers are defined in the Data Section and are used throughout the file to reference and identify specific building elements or attributes.

[0065] 2. Find geometric and location information:

[0066] Geometric information and location information are usually contained in the specific entity description. For example, for an entity representing a wall, its attribute part may include geometric shape (such as dimensions, coordinates), location information (such as relationships with other components), etc.

[0067] 3. Parse geometric shape information:

[0068] Geometric shape information may involve coordinate data, dimensions, shape descriptions, etc. This information is defined according to the IFC standard and corresponding descriptions can be found in the entity attributes of the STEP file.

[0069] In this way, the extraction of entity numbers, location information, and shape information corresponding to each entity number is achieved, as the shape modeling data of the target entity.

[0070] As the most common model data representation format at present, the STEP file can even be directly used in the BIM system. In this embodiment, the STEP file is used as a bridge. On the one hand, the universality of STEP is utilized to make this embodiment easier to implement. At the same time, the clear structure of the STEP file is utilized to achieve the extraction of shape modeling data.

[0071] Further, in a preferred embodiment, the steps performed by the above-mentioned model building module 120 are as follows: establishing a three-dimensional grid space and modeling based on the shape modeling data in the three-dimensional grid space to obtain a target entity model in the three-dimensional grid space, specifically including:

[0072] Establish a three-dimensional grid space and obtain the size data of the three-dimensional grid space;

[0073] Based on the size data of the three-dimensional grid space, the position information and shape information in the shape modeling data, perform coordinate transformation on the position information to obtain standard position data;

[0074] Based on the size data of the three-dimensional grid space, the position information and shape information in the shape modeling data, perform scaling on the shape information to obtain standard shape data;

[0075] Based on the entity number, the standard position information and standard shape information corresponding to each entity number, perform modeling in the three-dimensional grid space to obtain a target entity model in the three-dimensional grid space.

[0076] The above process corrects the shape modeling data based on the three-dimensional grid space through two means of coordinate transformation and size scaling to obtain unified standard position information and standard shape information. In this way, entities of any specification can be encoded through the three-dimensional grid space, and thus can be recognized through a preset three-dimensional convolutional neural network model, and only one preset three-dimensional convolutional neural network model needs to be trained.

[0077] After the target entity model is established, encoding can be performed according to the coverage relationship between the three-dimensional grid space and the target entity model. For example, if there is a target entity model in a three-dimensional grid, then the element value corresponding to the three-dimensional grid in the three-dimensional input matrix can be 1, otherwise it is 0.

[0078] Obviously, the above method will lose the information on the surface of the target entity model. For example, if a certain part on the edge of the target entity model does not fill the three-dimensional grid, setting the element value corresponding to the three-dimensional grid to 1 will result in a shape error.

[0079] Therefore, further, in a preferred embodiment, the steps performed by the above-mentioned data encoding module 130 are as follows: encoding according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix, specifically including:

[0080] Obtain a target grid, where the target grid is a grid in the three-dimensional grid space;

[0081] Calculate the space occupancy ratio of the target entity model in the target grid and normalize the space occupancy ratio to obtain the element value of the three-dimensional input matrix corresponding to the target grid.

[0082] The above design further describes the edge features of the target entity model by the spatial occupancy ratio of the target entity model, so as to retain more shape information in the three-dimensional input matrix, make the described features more accurate, and thus improve the accuracy of the output of the preset three-dimensional convolutional neural network.

[0083] Further, in a preferred embodiment, in the above-mentioned category recognition module 140, the specific structure of the preset three-dimensional convolutional neural network includes: an input layer, at least one three-dimensional convolutional layer, at least one pooling layer, at least one fully connected layer, and an output layer. The input layer is located at the first layer of the preset three-dimensional convolutional neural network. A three-dimensional convolutional layer is connected after the input layer. The output layer is located at the last layer of the preset three-dimensional convolutional neural network. A fully connected layer is connected before the output layer. The other three-dimensional convolutional layers, pooling layers, and fully connected layers are sequentially connected between the three-dimensional convolutional layer after the input layer and the fully connected layer before the output layer in an arranged and combined manner.

[0084] Among them, the input layer and the output layer are respectively used for inputting and outputting data. The three-dimensional convolutional layer is used for feature extraction of the three-dimensional input matrix. The pooling layer is used to reduce the data magnitude to reduce the operation and training pressure of the preset three-dimensional convolutional neural network and improve the running speed at the same time. The fully connected layer can be understood as performing complex non-linear analysis on the features extracted by the three-dimensional convolutional layer to obtain the result. It can be understood that in the above structure, the other three-dimensional convolutional layers, pooling layers, and fully connected layers in the middle part can be flexibly combined and arranged according to specific situations.

[0085] The three-dimensional convolutional layer after the input layer is the first three-dimensional convolutional layer for feature extraction, and its accuracy affects the accuracy of subsequent processing. It can perform convolution using a preset three-dimensional convolution kernel to achieve feature extraction. The present invention also provides a preferred embodiment for improving the first three-dimensional convolutional layer to obtain a more accurate feature extraction effect.

[0086] Specifically, in a preferred embodiment, the three-dimensional convolutional layer after the input layer includes a shape three-dimensional convolutional layer, a long-edge convolutional layer, a wide-edge convolutional layer, a high-edge convolutional layer, and a fusion layer. The input ends of the shape three-dimensional convolutional layer, the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer are all connected to the input layer. The output ends of the shape three-dimensional convolutional layer, the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer are all connected to the input end of the fusion layer. Among them, the shape three-dimensional convolutional layer is used to detect the shape features of the target entity model. The long-edge convolutional layer is used to detect the shape edge features of the target entity model in the long dimension. The wide-edge convolutional layer is used to detect the shape edge features of the target entity model in the wide dimension. The high-edge convolutional layer is used to detect the shape edge features of the target entity model in the high dimension. The fusion layer is used to respectively superimpose and fuse the output results of the long-edge convolutional layer, the wide-edge convolutional layer, and the high-edge convolutional layer into the output result of the shape three-dimensional convolutional layer in the three dimensions of length, width, and height, and output the first convolutional three-dimensional matrix.

[0087] It can be understood that the three dimensions of length, width, and height in the above process can be understood as three orthogonal directions in the three-dimensional grid space, or as the dimensions in which three numbers are arranged in the three-dimensional input matrix. Since in the previous encoding of the three-dimensional input matrix, the data at the model edge was processed more precisely through the spatial occupancy of the target entity model, in order to amplify this precision, in this embodiment, the first three-dimensional convolutional layer is further divided into a shape three-dimensional convolutional layer mainly used for extracting the shape, and a long-edge convolutional layer, a wide-edge convolutional layer, and a high-edge convolutional layer respectively used for extracting the edge features in the three dimensions. This design can be understood as further correcting the extracted shape features through the edge information of the three views of the target entity model. This kind of design can greatly increase the effective information volume of feature extraction and improve the accuracy of the entire preset three-dimensional convolutional neural network.

[0088] The above-mentioned shape three-dimensional convolutional layer, long-edge convolutional layer, wide-edge convolutional layer, and high-edge convolutional layer can all use the same specification of convolutional kernels (but the specific values of the convolutional kernels are different) to perform convolution on the three-dimensional input matrix, so as to ensure that the shape three-dimensional convolutional layer, long-edge convolutional layer, wide-edge convolutional layer, and high-edge convolutional layer all output data of the same specification (i.e., three-dimensional matrices of the same specification), so as to ensure that the subsequent fusion layer can perform fusion and superposition.

[0089] However, obviously, the above method will greatly increase the data volume of the first three-dimensional convolutional layer, so this embodiment makes an improvement on the above method again.

[0090] Specifically, in a preferred embodiment, the three-dimensional grid space has a size of n in each of the three dimensions of length, width, and height; the convolutional kernel of the shape three-dimensional convolutional layer has a size of a in each of the three dimensions of length, width, and height, where both a and n are positive integers and a is less than n. The convolutional kernel of the shape three-dimensional convolutional layer is used to move step by step in each of the three dimensions of the input three-dimensional matrix. The shape three-dimensional convolutional layer is used to output a shape feature three-dimensional matrix; the convolutional kernel of the long-edge convolutional layer has dimensions of n, a, and a in the three dimensions of length, width, and height respectively. The convolutional kernel of the long-edge convolutional layer is used to move step by step in each of the two dimensions of width and height of the input three-dimensional matrix. The long-edge convolutional layer is used to output a long-edge feature two-dimensional matrix; the convolutional kernel of the wide-edge convolutional layer has dimensions of a, n, and a in the three dimensions of length, width, and height respectively. The convolutional kernel of the wide-edge convolutional layer is used to move step by step in each of the two dimensions of length and height of the input three-dimensional matrix. The long-edge convolutional layer is used to output a wide-edge feature two-dimensional matrix; the convolutional kernel of the high-edge convolutional layer has dimensions of a, a, and n in the three dimensions of length, width, and height respectively. The convolutional kernel of the high-edge convolutional layer is used to move step by step in each of the two dimensions of length and width of the input three-dimensional matrix. The high-edge convolutional layer is used to output a high-edge feature two-dimensional matrix.

[0091] The above design enables the convolutional kernels in the long-edge convolutional layer, wide-edge convolutional layer, and high-edge convolutional layer to occupy the dimensions of the three-dimensional input matrix in one dimension respectively. In this way, the long-edge convolutional layer, wide-edge convolutional layer, and high-edge convolutional layer actually perform two-dimensional convolution, greatly reducing the data volume in the preset three-dimensional convolutional neural network. And because the long-edge convolutional layer, wide-edge convolutional layer, and high-edge convolutional layer are themselves used to extract edge features, this design actually does not affect the accuracy of overall feature extraction. While simplifying the network structure, it can also ensure accuracy. Figure 2 It is a schematic structural diagram of a simplified first three-dimensional convolutional layer.

[0092] Further, in a preferred embodiment, the fusion layer is used to stack and fuse the long-edge feature two-dimensional matrix, wide-edge feature two-dimensional matrix, and high-edge feature two-dimensional matrix into the shape feature three-dimensional matrix in each of the three dimensions of length, width, and height, and output a first convolutional three-dimensional matrix. The corresponding formula is:

[0093]

[0094] where represents the value of the element at coordinates i, j, and k in the three dimensions of length, width, and height in the first convolutional three-dimensional matrix, represents the value of the element at coordinates i, j, and k in the three dimensions of length, width, and height in the shape feature three-dimensional matrix, Denotes the value of the element at coordinates j and k in the two-dimensional matrix of long-edge features in the width and height dimensions respectively, Denotes the value of the element at coordinates i and k in the two-dimensional matrix of wide-edge features in the length and height dimensions respectively, Denotes the value of the element at coordinates i and j in the two-dimensional matrix of high-edge features in the length and width dimensions respectively, and w0 represents the element in the fusion layer The weight value between the corresponding node and the element in the three-dimensional shape convolution layer The weight value between the corresponding node and the element in the long-edge convolution layer, where w1 represents the element in the fusion layer The weight value between the corresponding node and the element in the wide-edge convolution layer The weight value between the corresponding node and the element in the wide-edge convolution layer, where w2 represents the element in the fusion layer The weight value between the corresponding node and the element in the wide-edge convolution layer The weight value between the corresponding node and the element in the wide-edge convolution layer, where w3 represents the element in the fusion layer The weight value between the corresponding node and the element in the high-edge convolution layer The weight value between the corresponding node and the element in the high-edge convolution layer, b i,j,k Denotes the bias value of the element in the fusion layer The corresponding node, f() is a preset activation function, and m is the number of elements in the length dimension of the three-dimensional shape feature matrix.

[0095] In practice, the two-dimensional matrix of long-edge features, the two-dimensional matrix of wide-edge features, and the two-dimensional matrix of high-edge features can be directly fused into the three-dimensional shape feature matrix by simple linear superposition. In this embodiment, the fusion layer is directly connected to the three-dimensional shape convolution layer, the long-edge convolution layer, the wide-edge convolution layer, and the high-edge convolution layer through a fully connected manner, so that the fusion rule of the fusion layer can be trained (i.e., the specific values of w0, w1, w2, w3, and b i,j,k in the above formula), further improving the accuracy of feature extraction.

[0096] Furthermore, in a preferred embodiment, in the preset three-dimensional convolutional neural network, the output layer includes multiple output nodes, and the multiple output nodes are respectively connected to multiple nodes in the fully connected layer adjacent to the output layer. The number of output nodes is equal to the preset number of types of target entities, and each output node corresponds to a preset type. The output node is used to output the probability value that the type of the target entity is its corresponding preset type.

[0097] This structure has high interpretability, enabling the system to classify in an efficient and accurate manner by analyzing the shape information of the target entity. Through the training and optimization of the three-dimensional convolutional neural network, the ability to automatically identify different entity types can be improved, thereby accelerating the speed of modeling and data entry, reducing the need for manual intervention, and enhancing the overall work efficiency and accuracy.

[0098] Further, in a preferred embodiment, the building information includes material information, attribute information, cost information, spatial information, and structural information. The steps performed by the above-mentioned information supplement module 150 are as follows: Based on the target entity type, supplement the building information of the target entity in the BIM system, specifically including:

[0099] Retrieve entities of the same type as the target entity type in the BIM system;

[0100] Obtain the material information of the entities of the same type as the target entity type in the BIM system as the material information of the target entity;

[0101] Based on the material information of the target entity, supplement the attribute information, cost information, spatial information, and structural information of the target entity.

[0102] The above process uses the known information of entities of the same type in the BIM system to supplement data according to the target entity type, improving the accuracy of data filling. At the same time, using the material information as the basis, most of the building data can be supplemented. For example, after knowing the shape information and material information of the target entity in the building, the attribute information of the target entity (such as material properties of building components, such as type, strength, durability, environmental protection performance, etc., and technical properties such as heat insulation performance, fire protection performance, and acoustic performance), cost information (such as material budgets and material expenditures of each component, system, or the entire project), spatial information (functional zoning and usage planning of the internal space of the building, actual usage and occupancy of each space, etc.), and structural information (such as structural calculation and analysis results, including load distribution, stress and strain, etc., or safety performance such as seismic performance and fire protection performance of the structure) can be calculated. The above information can be inferred using any existing technology, so no further description is provided in this article.

[0103] The present invention provides a building information management system based on BIM technology. Through a data parsing module, shape modeling data of a target entity is obtained based on a BIM model. Then, a three-dimensional grid space is established by a model building module and modeling is carried out in the three-dimensional grid space to obtain a target entity model. Next, a data encoding module encodes according to the coverage relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix. After that, a type recognition module uses the three-dimensional input matrix and a preset three-dimensional convolutional neural network to identify the type of the target entity. Finally, an information supplement module supplements the building information of the target entity in the BIM system based on the type of the target entity. Compared with the prior art, the present invention is based on the most important and fundamental shape information in the BIM system, that is, the shape modeling data, conducts modeling in the three-dimensional grid space, and uses the coverage relationship between the three-dimensional grid space and the target entity model to complete the encoding of the entity shape. In this way, the type of the entity can be identified through the three-dimensional convolutional neural network, and then other missing information of the target entity can be automatically supplemented based on the type of the target entity. In this way, the entire BIM system can temporarily start working based on the automatically filled building information and make changes when subsequent information is entered, solving the problem that the BIM technology in the prior art cannot be normally applied in the case of data loss.

[0104] It should be noted that the embodiments in this specification are all described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other.

[0105] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A building information management system based on BIM technology, characterized in that: include: A data analysis module is used to obtain shape modeling data of a target entity based on a BIM model; A model building module is used to build a three-dimensional grid space and to build a model in the three-dimensional grid space based on the shape modeling data to obtain a target entity model in the three-dimensional grid space; A data encoding module is used to encode according to the covering relationship between the three-dimensional grid space and the target entity model to obtain a three-dimensional input matrix, wherein each element in the three-dimensional input matrix corresponds to a grid in the three-dimensional grid space, and the value of each element in the three-dimensional input matrix is ​​used to represent a part of the target entity model in the corresponding grid; A category recognition module, used to input the three-dimensional input matrix into a preset three-dimensional convolutional neural network to obtain the target entity category output by the preset three-dimensional convolutional neural network; An information supplement module is used to supplement the building information of the target entity in the BIM system based on the target entity type; Among them, based on the BIM model, the shape modeling data of the target entity is obtained, including: Obtain input range selection data representing the target entity, and obtain BIM shape data of the target entity according to the BIM model based on the range selection data; Based on the BIM shape data, obtain the general model data of the target entity in STEP format; The entity number in the general model data, the position information and shape information corresponding to each entity number are obtained as the shape modeling data of the target entity.

2. The building information management system based on BIM technology according to claim 1 is characterized in that: A three-dimensional grid space is established, and modeling is performed in the three-dimensional grid space based on the shape modeling data to obtain a target entity model in the three-dimensional grid space, including: Establishing a three-dimensional grid space and obtaining dimension data of the three-dimensional grid space; Based on the size data of the three-dimensional grid space, the position information and the shape information in the shape modeling data, coordinate transformation is performed on the position information to obtain standard position information; Based on the size data of the three-dimensional grid space, the position information and the shape information in the shape modeling data, the shape information is scaled to obtain the standard shape information; Based on the entity number, the standard position information and the standard shape information corresponding to each entity number, modeling is performed in the three-dimensional grid space to obtain a target entity model in the three-dimensional grid space.

3. The building information management system based on BIM technology according to claim 2 is characterized in that: According to the covering relationship between the three-dimensional grid space and the target entity model, the three-dimensional input matrix is ​​obtained, including: Obtain a target grid, where the target grid is a grid in a three-dimensional grid space; The spatial proportion of the target entity model in the target grid is calculated and normalized to obtain the element value of the three-dimensional input matrix corresponding to the target grid.

4. The building information management system based on BIM technology according to claim 1 is characterized in that: The preset three-dimensional convolutional neural network includes an input layer, at least one three-dimensional convolutional layer, at least one pooling layer, at least one fully connected layer and an output layer, wherein the input layer is located at the first layer of the preset three-dimensional convolutional neural network, a three-dimensional convolutional layer is connected after the input layer, the output layer is located at the last layer of the preset three-dimensional convolutional neural network, a fully connected layer is connected before the output layer, and other three-dimensional convolutional layers, pooling layers and fully connected layers are arranged and combined and connected in sequence between the three-dimensional convolutional layer after the input layer and the fully connected layer before the output layer.

5. The building information management system based on BIM technology according to claim 4 is characterized in that: The three-dimensional convolution layer after the input layer includes a shape three-dimensional convolution layer, a long edge convolution layer, a wide edge convolution layer, a high edge convolution layer and a fusion layer, wherein the input ends of the shape three-dimensional convolution layer, the long edge convolution layer, the wide edge convolution layer and the high edge convolution layer are all connected to the input layer, and the output ends of the shape three-dimensional convolution layer, the long edge convolution layer, the wide edge convolution layer and the high edge convolution layer are all connected to the input end of the fusion layer, wherein the shape three-dimensional convolution layer is used to detect the shape features of the target entity model, the long edge convolution layer is used to detect the shape edge features of the target entity model in the long dimension, the wide edge convolution layer is used to detect the shape edge features of the target entity model in the wide dimension, and the high edge convolution layer is used to detect the shape edge features of the target entity model in the high dimension, and the fusion layer is used to superimpose and fuse the output results of the long edge convolution layer, the wide edge convolution layer and the high edge convolution layer into the output results of the shape three-dimensional convolution layer in the three dimensions of length, width and height, and output the first convolution three-dimensional matrix.

6. The building information management system based on BIM technology according to claim 5 is characterized in that: The size of the three-dimensional grid space in length, width and height is n; the size of the convolution kernel in the three dimensions of length, width and height of the shape three-dimensional convolution layer is a, where a and n are both positive integers and a is less than n. The convolution kernel of the shape three-dimensional convolution layer is used to move step by step in the length, width and height of the input three-dimensional matrix respectively, and the shape three-dimensional convolution layer is used to output a three-dimensional matrix of shape features; The convolution kernel of the long edge convolution layer has the sizes of n, a and a in length, width and height respectively. The convolution kernel of the long edge convolution layer is used to move step by step in the width and height dimensions of the input three-dimensional matrix respectively. The long edge convolution layer is used to output a two-dimensional matrix of long edge features. The convolution kernel of the wide edge convolution layer has the sizes of a, n and a in length, width and height respectively. The convolution kernel of the wide edge convolution layer is used to move step by step in the length and height of the input three-dimensional matrix respectively. The wide edge convolution layer is used to output a two-dimensional matrix of wide edge features. The sizes of the convolution kernel of the high edge convolution layer in length, width and height are a, a and n respectively. The convolution kernel of the high edge convolution layer is used to move step by step in the length and width of the input three-dimensional matrix respectively. The high edge convolution layer is used to output a high edge feature two-dimensional matrix.

7. The building information management system based on BIM technology according to claim 6 is characterized in that: The fusion layer is used to superimpose and fuse the long edge feature two-dimensional matrix, the wide edge feature two-dimensional matrix and the high edge feature two-dimensional matrix into the shape feature three-dimensional matrix in the three dimensions of length, width and height, and output the first convolution three-dimensional matrix. The corresponding formula is: in, Represents the values ​​of the elements whose coordinates in length, width and height are i, j and k in the first convolution three-dimensional matrix, Represents the values ​​of the elements whose coordinates in the three dimensions of length, width and height in the three-dimensional matrix of shape features are i, j and k respectively. Represents the value of the element with coordinates j and k in the width and height dimensions in the two-dimensional matrix of the long edge feature, Represents the value of the element whose coordinates in the length and height dimensions are i and k in the two-dimensional matrix of the wide edge feature. Represents the value of the element with coordinates i and j in the length and width dimensions in the two-dimensional matrix of the high edge feature, and w0 represents the element in the fusion layer Corresponding nodes and shapes of elements in the 3D convolutional layer The weight values ​​between the corresponding nodes, w1 represents the elements in the fusion layer The corresponding nodes and elements in the long edge convolution layer The weight value between the corresponding nodes, w2 represents the element in the fusion layer The corresponding nodes and elements in the wide edge convolution layer The weight values ​​between the corresponding nodes, w3 represents the elements in the fusion layer The corresponding nodes and elements in the high edge convolution layer The corresponding weight value between nodes, b i,j,k Represents the elements in the fusion layer The corresponding node bias value, f() is the preset activation function, and m is the number of elements in the long dimension of the shape feature three-dimensional matrix.

8. The building information management system based on BIM technology according to claim 4 is characterized in that: The output layer includes multiple output nodes, and the multiple output nodes are respectively connected to multiple nodes in the fully connected layer adjacent to the output layer. The number of output nodes is equal to the preset number of target entity types. Each output node corresponds to a preset type, and the output node is used to output the probability value of the target entity type being the corresponding preset type.

9. The building information management system based on BIM technology according to claim 1, characterized in that: Building information includes material information, attribute information, cost information, space information and structural information; Based on the target entity type, the building information of the target entity is supplemented in the BIM system, including: Retrieve entities of the same type as the target entity in the BIM system; Obtain material information of entities of the same type as the target entity in the BIM system as the material information of the target entity; Based on the material information of the target entity, the attribute information, cost information, space information and structure information of the target entity are supplemented; Among them, based on the BIM model, the shape modeling data of the target entity is obtained, including: Obtain input range selection data representing the target entity, and obtain BIM shape data of the target entity according to the BIM model based on the range selection data; Based on the BIM shape data, obtain the general model data of the target entity in STEP format; The entity number in the general model data, the position information and shape information corresponding to each entity number are obtained as the shape modeling data of the target entity.

Images