BIM model split display method, device, equipment and storage medium

By acquiring and analyzing the functional area, system composition and spatial structure information of the BIM model, and combining the model geometry separation network and boundary interaction analysis, the model splitting and display are optimized, and the problem of low splitting accuracy of complex BIM models is solved. Achieve more efficient display effect.

CN117407959BActive Publication Date: 2024-12-06CHANGSHA JUNYAN TECH CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202311420132.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-10-30
Publication Date
2024-12-06
Estimated Expiration
2043-10-30

AI Technical Summary

Technical Problem

The prior art is difficult to accurately split and display complex large-scale BIM models, resulting in low spin-off accuracy.

Method used

By obtaining the functional area information, system composition information and spatial structure information of the target building BIM model, feature extraction and analysis are carried out, a collection of target selection factors is generated, and inputting it to the model geometry separation network for model geometry splitting, combining with the boundary interaction analysis model and boundary relationship matrix, optimize model boundary and connection relationships, and finally render and display.

Benefits of technology

The accuracy of spin-off and display accuracy of the BIM model is improved, making the splitting results of the building more intuitive and visual.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117407959B_ABST
    Figure CN117407959B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of artificial intelligence, and discloses a method, device, equipment and storage medium for splitting and displaying a BIM model, which are used to improve the splitting accuracy of the BIM model and improve the display accuracy of the split BIM model. The method includes: inputting a first building model splitting scheme and a target building BIM model into a preset model geometry separation network to split the model geometry, and obtain multiple first split BIM models; extracting the model boundary to obtain multiple model boundary features and performing connection relationship analysis to obtain multiple connection relationship features; constructing a target boundary relationship matrix, and inputting the target boundary relationship matrix into a boundary interaction analysis model to perform boundary interaction analysis to generate a second building model splitting scheme; optimizing the model boundary and connection relationship according to the second building model splitting scheme to generate multiple second split BIM models, and rendering and displaying the multiple second split BIM models.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a method, device, equipment and storage medium for disassembling and displaying a BIM model. Background Art

[0002] BIM Building Information Model is a digital tool widely used in the process of building design and construction. It provides full life cycle management of building projects by integrating the geometry, attributes and relationship data of building design into one model. However, as the complexity of construction projects increases, traditional BIM models tend to become large and difficult to handle. Therefore, researchers have been looking for effective methods to split, display and optimize large BIM models.

[0003] However, the existing solutions rely on manual intervention by users. For complex building structures and large BIM models, it is difficult to separate accurate BIM models based on manual experience, that is, the accuracy of the existing solutions is low. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for splitting and displaying a BIM model, which are used to improve the splitting accuracy of the BIM model and improve the display accuracy of the split BIM model.

[0005] A first aspect of the present invention provides a method for disassembling and displaying a BIM model, the method comprising:

[0006] Acquire functional area information and system composition information of the target building BIM model, and perform spatial structure analysis on the target building BIM model to obtain spatial structure information;

[0007] Performing feature extraction on the functional area information, the system composition information, and the spatial structure information to obtain a target selection factor set, and creating a first building model splitting scheme according to the target selection factor set;

[0008] Inputting the first building model splitting scheme and the target building BIM model into a preset model geometry separation network to perform model geometry splitting to obtain a plurality of first split BIM models;

[0009] Performing model boundary extraction on the multiple first split BIM models to obtain multiple model boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple connection relationship features;

[0010] Constructing a target boundary relationship matrix according to the plurality of model boundary features and the plurality of connection relationship features, and inputting the target boundary relationship matrix into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting solution;

[0011] According to the second building model splitting scheme, the model boundaries and connection relationships of the multiple first split BIM models are optimized to generate multiple second split BIM models, and the multiple second split BIM models are rendered and displayed.

[0012] In combination with the first aspect, in a first implementation of the first aspect of the present invention, the acquiring of functional area information and system composition information of a target building BIM model, and performing spatial structure analysis on the target building BIM model to obtain spatial structure information includes:

[0013] Obtain a model configuration file of the target building BIM model, and perform model configuration extraction on the model configuration file to obtain model configuration information to be matched;

[0014] Performing functional feature matching on the to-be-matched model configuration information to obtain functional area information, and performing system composition matching on the to-be-matched model configuration information to obtain system composition information;

[0015] Performing spatial distribution analysis on the target building BIM model to obtain an initial spatial distribution, and performing spatial classification on the target building BIM model to obtain a target spatial type;

[0016] Determining a corresponding target space distribution according to the target space type, and performing a collision detection on the initial space distribution and the target space distribution to obtain a collision detection result;

[0017] According to the collision detection result, a hierarchical structure analysis is performed on the target spatial distribution and the initial spatial distribution to generate spatial structure information of the target building BIM model.

[0018] In combination with the first aspect, in a second implementation of the first aspect of the present invention, the feature extraction of the functional area information, the system composition information, and the spatial structure information to obtain a target selection factor set, and creating a first building model splitting scheme according to the target selection factor set includes:

[0019] Performing feature analysis on the functional area information to obtain a plurality of functional area features, wherein the plurality of functional area features include: area, function and attribute of the area;

[0020] Extracting features from the system composition information to obtain a plurality of system composition features, wherein the plurality of system composition features include: water supply system features, air conditioning system features, and power system features;

[0021] Performing feature clustering on the spatial structure information to obtain a plurality of spatial structure features, wherein the plurality of spatial structure features include: floor division, room layout, and spatial connection relationship;

[0022] Performing selection factor rule matching on the multiple functional area features, the multiple system composition features, and the multiple spatial structure features to obtain multiple selection factor rules;

[0023] Creating a set through the multiple selection factor rules to obtain a target selection factor set;

[0024] By using the target selection factor set, a sub-building model splitting scheme corresponding to each selection factor rule is created;

[0025] A first building model splitting scheme is generated according to the sub-building model splitting scheme corresponding to each selection factor rule.

[0026] In combination with the first aspect, in a third implementation of the first aspect of the present invention, the first building model splitting scheme and the target building BIM model are input into a preset model geometry separation network for model geometry splitting to obtain a plurality of first split BIM models, including:

[0027] Performing geometric transformation on the target building BIM model to obtain target geometric data, and inputting the target geometric data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network includes: a feature extraction layer, a separation layer, and an output layer;

[0028] Performing feature extraction on the first building model splitting scheme through the feature extraction layer to obtain a plurality of building model splitting virtual areas and virtual area feature information corresponding to each building model splitting virtual area;

[0029] Through the separation layer, the target geometric body data is geometrically separated according to the multiple building model split virtual areas and virtual area feature information corresponding to each building model split virtual area to obtain multiple split geometric bodies;

[0030] Through the output layer, geometric tags of the multiple split geometric bodies are created, and the multiple split geometric bodies are output as corresponding multiple first split BIM models according to the geometric tags.

[0031] In combination with the first aspect, in a fourth implementation of the first aspect of the present invention, the performing of model boundary extraction on the multiple first split BIM models to obtain multiple model boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple connection relationship features include:

[0032] Performing model boundary extraction on the multiple first split BIM models to obtain multiple original boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple original relationship features;

[0033] Inputting the plurality of original boundary features into a preset feature clustering model, and calculating cluster centers of the plurality of original boundary features by using a preset first feature clustering function to obtain a first initial cluster center, inputting the plurality of original relationship features into the feature clustering model, and calculating cluster centers of the plurality of original relationship features by using a preset second feature clustering function to obtain a second initial cluster center;

[0034] Performing distance calculation on the multiple original boundary features and the first initial cluster center to obtain multiple first cluster distances and performing mean operation on the multiple first cluster distances to obtain a first average distance, and performing distance calculation on the multiple original relationship features and the second initial cluster center to obtain multiple second cluster distances and performing mean operation on the multiple second cluster distances to obtain a second average distance;

[0035] Adjusting the parameters of the first feature clustering function according to the first average distance to obtain a first target clustering function, and adjusting the parameters of the second feature clustering function according to the second average distance to obtain a second target clustering function;

[0036] Determining a first target cluster center according to the first target cluster function, and determining a second target cluster center according to the second target cluster function;

[0037] The plurality of original boundary features are clustered again using the first target cluster center to obtain a plurality of model boundary features, and the plurality of original relationship features are clustered again using the second target cluster center to obtain a plurality of connection relationship features.

[0038] In combination with the first aspect, in a fifth implementation of the first aspect of the present invention, constructing a target boundary relationship matrix according to the multiple model boundary features and the multiple connection relationship features, and inputting the target boundary relationship matrix into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting solution, including:

[0039] Performing feature value mapping on the multiple model boundary features and the multiple connection relationship features to obtain multiple boundary feature values ​​and multiple relationship feature values;

[0040] Performing matrix conversion on the plurality of boundary eigenvalues ​​and the plurality of relationship eigenvalues ​​to obtain a target boundary relationship matrix;

[0041] Inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model includes: a first encoding network, a second encoding network and a decoding network;

[0042] Performing boundary interaction analysis on the target boundary relationship matrix through the boundary interaction analysis model to generate a target prediction probability;

[0043] According to the target prediction probability, a corresponding second building model splitting scheme is matched from a plurality of preset candidate building model splitting schemes.

[0044] In combination with the first aspect, in a sixth implementation of the first aspect of the present invention, optimizing the model boundaries and connection relationships of the multiple first split BIM models according to the second building model splitting scheme, generating multiple second split BIM models, and rendering and displaying the multiple second split BIM models include:

[0045] Reorganizing the plurality of first split BIM models to obtain a reorganized building BIM model;

[0046] According to the second building model splitting scheme, adjusting model boundary parameters and optimizing connection relationships of the reorganized building BIM model to generate a plurality of second split BIM models;

[0047] The multiple second split BIM models are rendered to obtain multiple rendered BIM models, and the multiple rendered BIM models are displayed in virtual reality.

[0048] A second aspect of the present invention provides a disassembly and display device for a BIM model, the disassembly and display device for a BIM model comprising:

[0049] An acquisition module is used to acquire functional area information and system composition information of a target building BIM model, and to perform spatial structure analysis on the target building BIM model to obtain spatial structure information;

[0050] A creation module, used for performing feature extraction on the functional area information, the system composition information and the spatial structure information to obtain a target selection factor set, and creating a first building model splitting scheme according to the target selection factor set;

[0051] A splitting module, used for inputting the first building model splitting scheme and the target building BIM model into a preset model geometry separation network to perform model geometry splitting to obtain a plurality of first split BIM models;

[0052] A processing module, configured to extract model boundaries from the plurality of first split BIM models to obtain a plurality of model boundary features, and perform connection relationship analysis on the plurality of first split BIM models to obtain a plurality of connection relationship features;

[0053] An analysis module, configured to construct a target boundary relationship matrix according to the plurality of model boundary features and the plurality of connection relationship features, and input the target boundary relationship matrix into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting solution;

[0054] A display module is used to optimize the model boundaries and connection relationships of the multiple first split BIM models according to the second building model splitting scheme, generate multiple second split BIM models, and render and display the multiple second split BIM models.

[0055] In combination with the second aspect, in a first implementation of the second aspect of the present invention, the acquisition module is specifically used to:

[0056] Obtain a model configuration file of the target building BIM model, and perform model configuration extraction on the model configuration file to obtain model configuration information to be matched;

[0057] Performing functional feature matching on the to-be-matched model configuration information to obtain functional area information, and performing system composition matching on the to-be-matched model configuration information to obtain system composition information;

[0058] Performing spatial distribution analysis on the target building BIM model to obtain an initial spatial distribution, and performing spatial classification on the target building BIM model to obtain a target spatial type;

[0059] Determining a corresponding target space distribution according to the target space type, and performing a collision detection on the initial space distribution and the target space distribution to obtain a collision detection result;

[0060] According to the collision detection result, a hierarchical structure analysis is performed on the target spatial distribution and the initial spatial distribution to generate spatial structure information of the target building BIM model.

[0061] In combination with the second aspect, in a second implementation of the second aspect of the present invention, the creation module is specifically used to:

[0062] Performing feature analysis on the functional area information to obtain a plurality of functional area features, wherein the plurality of functional area features include: area, function and attribute of the area;

[0063] Extracting features from the system composition information to obtain a plurality of system composition features, wherein the plurality of system composition features include: water supply system features, air conditioning system features, and power system features;

[0064] Performing feature clustering on the spatial structure information to obtain a plurality of spatial structure features, wherein the plurality of spatial structure features include: floor division, room layout, and spatial connection relationship;

[0065] Performing selection factor rule matching on the multiple functional area features, the multiple system composition features, and the multiple spatial structure features to obtain multiple selection factor rules;

[0066] Creating a set through the multiple selection factor rules to obtain a target selection factor set;

[0067] By using the target selection factor set, a sub-building model splitting scheme corresponding to each selection factor rule is created;

[0068] A first building model splitting scheme is generated according to the sub-building model splitting scheme corresponding to each selection factor rule.

[0069] In combination with the second aspect, in a third implementation of the second aspect of the present invention, the splitting module is specifically used for:

[0070] Performing geometric transformation on the target building BIM model to obtain target geometric data, and inputting the target geometric data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network includes: a feature extraction layer, a separation layer, and an output layer;

[0071] Performing feature extraction on the first building model splitting scheme through the feature extraction layer to obtain a plurality of building model splitting virtual areas and virtual area feature information corresponding to each building model splitting virtual area;

[0072] Through the separation layer, the target geometric body data is geometrically separated according to the multiple building model split virtual areas and virtual area feature information corresponding to each building model split virtual area to obtain multiple split geometric bodies;

[0073] Through the output layer, geometric tags of the multiple split geometric bodies are created, and the multiple split geometric bodies are output as corresponding multiple first split BIM models according to the geometric tags.

[0074] In combination with the second aspect, in a fourth implementation of the second aspect of the present invention, the processing module is specifically used to:

[0075] Performing model boundary extraction on the multiple first split BIM models to obtain multiple original boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple original relationship features;

[0076] Inputting the plurality of original boundary features into a preset feature clustering model, and calculating cluster centers of the plurality of original boundary features by using a preset first feature clustering function to obtain a first initial cluster center, inputting the plurality of original relationship features into the feature clustering model, and calculating cluster centers of the plurality of original relationship features by using a preset second feature clustering function to obtain a second initial cluster center;

[0077] Performing distance calculation on the multiple original boundary features and the first initial cluster center to obtain multiple first cluster distances and performing mean operation on the multiple first cluster distances to obtain a first average distance, and performing distance calculation on the multiple original relationship features and the second initial cluster center to obtain multiple second cluster distances and performing mean operation on the multiple second cluster distances to obtain a second average distance;

[0078] Adjusting the parameters of the first feature clustering function according to the first average distance to obtain a first target clustering function, and adjusting the parameters of the second feature clustering function according to the second average distance to obtain a second target clustering function;

[0079] Determining a first target cluster center according to the first target cluster function, and determining a second target cluster center according to the second target cluster function;

[0080] The plurality of original boundary features are clustered again using the first target cluster center to obtain a plurality of model boundary features, and the plurality of original relationship features are clustered again using the second target cluster center to obtain a plurality of connection relationship features.

[0081] In combination with the second aspect, in a fifth implementation of the second aspect of the present invention, the analysis module is specifically used for:

[0082] Performing feature value mapping on the multiple model boundary features and the multiple connection relationship features to obtain multiple boundary feature values ​​and multiple relationship feature values;

[0083] Performing matrix conversion on the plurality of boundary eigenvalues ​​and the plurality of relationship eigenvalues ​​to obtain a target boundary relationship matrix;

[0084] Inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model includes: a first encoding network, a second encoding network and a decoding network;

[0085] Performing boundary interaction analysis on the target boundary relationship matrix through the boundary interaction analysis model to generate a target prediction probability;

[0086] According to the target prediction probability, a corresponding second building model splitting scheme is matched from a plurality of preset candidate building model splitting schemes.

[0087] In combination with the second aspect, in a sixth implementation of the second aspect of the present invention, the display module is specifically used for:

[0088] Reorganizing the plurality of first split BIM models to obtain a reorganized building BIM model;

[0089] According to the second building model splitting scheme, adjusting model boundary parameters and optimizing connection relationships of the reorganized building BIM model to generate a plurality of second split BIM models;

[0090] The multiple second split BIM models are rendered to obtain multiple rendered BIM models, and the multiple rendered BIM models are displayed in virtual reality.

[0091] A third aspect of the present invention provides a BIM model disassembly and display device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the BIM model disassembly and display device executes the above-mentioned BIM model disassembly and display method.

[0092] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for disassembling and displaying a BIM model.

[0093] In the technical solution provided by the present invention, the first building model splitting scheme and the target building BIM model are input into a preset model geometry separation network for model geometry splitting to obtain multiple first split BIM models; model boundary extraction is performed to obtain multiple model boundary features and connection relationship analysis is performed to obtain multiple connection relationship features; a target boundary relationship matrix is ​​constructed, and the target boundary relationship matrix is ​​input into a boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting scheme; model boundaries and connection relationships are optimized according to the second building model splitting scheme to generate multiple second split BIM models, and multiple second split BIM models are rendered and displayed. The present invention can generate a target selection factor set and a first building model splitting scheme by acquiring functional area information, system composition information and spatial structure information of the target building BIM model, and performing feature extraction and analysis. By inputting the first building model splitting scheme into a model geometry separation network for model geometry splitting, the method can generate multiple first split BIM models. Model boundary extraction and connection relationship analysis are performed on these models to further optimize the boundary and connection relationship of the model. By applying the boundary interaction analysis model and the target boundary relationship matrix, boundary interaction analysis can be performed and a second building model splitting scheme can be generated. By optimizing the model boundaries and connection relationships of multiple second split BIM models and applying rendering and display technology, the splitting results of the building can be understood and displayed more intuitively, thereby improving the splitting accuracy of the BIM model and the display accuracy of the split BIM model. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] Figure 1 A schematic diagram of an embodiment of a method for disassembling and displaying a BIM model in an embodiment of the present invention;

[0095] Figure 2 A flowchart of creating a first building model splitting solution in an embodiment of the present invention;

[0096] Figure 3 A flowchart of model geometry splitting in an embodiment of the present invention;

[0097] Figure 4 A flowchart of boundary interaction analysis in an embodiment of the present invention;

[0098] Figure 5 A schematic diagram of an embodiment of a disassembled display device of a BIM model in an embodiment of the present invention;

[0099] Figure 6 It is a schematic diagram of an embodiment of a disassembled display device of a BIM model in an embodiment of the present invention. DETAILED DESCRIPTION

[0100] The embodiments of the present invention provide a method, device, equipment and storage medium for displaying the splitting of a BIM model, which are used to improve the accuracy of the splitting of the BIM model and improve the display accuracy of the split BIM model. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0101] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the method for disassembling and displaying a BIM model in an embodiment of the present invention includes:

[0102] S101, obtaining functional area information and system composition information of a target building BIM model, and performing spatial structure analysis on the target building BIM model to obtain spatial structure information;

[0103] It is understandable that the execution subject of the present invention may be a disassembled display device of a BIM model, or may be a terminal or a server, which is not specifically limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0104] Specifically, it is necessary to obtain the model configuration file of the target building BIM model and extract the model configuration from the file. The model configuration file is a file that records various model elements, attributes and relationships in the BIM model. By parsing the model configuration file, the model configuration information to be matched can be extracted, including the type, attributes and relationships of the building elements. For the model configuration information to be matched, functional feature matching is performed to obtain functional area information. By matching spatial elements with similar functional attributes, they can be classified into the same functional area. For example, in an office building model, elements such as offices, conference rooms, corridors, etc. can be matched as different functional areas according to the attributes of spatial elements. At the same time, system composition matching is performed on the model configuration information to be matched to obtain system composition information. By matching elements with similar system attributes, they can be classified into the same system composition. For example, in a building model, matching can be performed according to the heating, ventilation and air conditioning system attributes of the elements, and elements with the same system attributes can be classified into one system composition. The spatial distribution analysis of the target building BIM model is performed to obtain the initial spatial distribution. This involves analyzing the position, relative distance, height and other information between spatial elements. By analyzing this information, the initial spatial distribution of the target building BIM model can be obtained. Based on the spatial distribution analysis, spatial classification is performed to obtain the target spatial type. According to the analysis results of the initial spatial distribution, the spatial elements can be classified, such as offices, conference rooms, corridors, etc. are classified into different spatial types. According to the target spatial type, the corresponding target spatial distribution is determined, and collision detection is performed on the initial spatial distribution and the target spatial distribution. The collision detection results can be obtained by detecting the overlap, intersection or gap between spatial elements. For example, in an office building model, if there is an overlap or unreasonable spatial relationship between offices, feedback will be obtained in the collision detection results. Based on the collision detection results, the target spatial distribution and the initial spatial distribution are hierarchically analyzed to generate the spatial structure information of the target building BIM model. By analyzing the parent-child relationship, nested relationship or context relationship between spatial elements, spatial structure information with a hierarchical structure can be constructed. For example, consider a BIM model of a residential building. By parsing the model configuration file, the model configuration information to be matched is extracted, including the types, attributes and relationships of spatial elements such as rooms, kitchens, living rooms, and bedrooms. Through functional feature matching, spatial elements with similar functional attributes are matched into different functional areas, such as taking the bedroom as a functional area. At the same time, through system composition matching, elements with the same heating system or air-conditioning system are classified into the same system composition. Then, the spatial distribution analysis of the building BIM model is performed to obtain the initial spatial distribution, such as the distribution of bedrooms upstairs and living rooms downstairs. According to the analysis results of the initial spatial distribution, the spatial elements are classified, such as classifying bedrooms, kitchens, and living rooms into different spatial types.According to the target space type, determine the target space distribution, such as setting the expected distribution of bedrooms on the upper floors of the building. Perform collision detection. If overlap is found between bedrooms, that is, a collision occurs, it will be fed back in the collision detection results. Based on the collision detection results, perform a hierarchical structure analysis on the target space distribution and the initial space distribution to generate the spatial structure information of the residential building BIM model, such as determining the parent-child relationship and nested relationship between the bedroom and the kitchen.

[0105] S102, extracting features from the functional area information, the system composition information, and the spatial structure information to obtain a target selection factor set, and creating a first building model splitting scheme according to the target selection factor set;

[0106] Specifically, feature analysis is performed on the functional area information to obtain multiple functional area features. These features may include area area, area function, and area attribute. By analyzing the attributes of the functional area, different functional area features can be identified, such as office area, meeting area, and rest area. Feature extraction is performed on the system composition information to obtain multiple system composition features. These features may include water supply system features, air conditioning system features, and power system features. By extracting the attributes and relationships of system elements, different system composition features can be identified, such as water supply pipes, air conditioning terminal equipment, and power distribution boards. At the same time, feature clustering is performed on the spatial structure information to obtain multiple spatial structure features. These features may include floor division, room layout, and spatial connection relationship. By analyzing the location, relationship, and function of spatial elements, different spatial structure features can be identified, such as the distribution of floors, the layout of rooms, and the connection mode of corridors. Selection factor rule matching is performed on multiple functional area features, system composition features, and spatial structure features to obtain multiple selection factor rules. The selection factor rule is a rule based on the feature extraction result, which is used to evaluate and select a suitable sub-building model splitting scheme. For example, a selection factor rule may be "If the functional area is an office area, the system composition includes a water supply system and an air conditioning system, and the spatial structure feature is that the floor is divided into an office area and a conference area, it is preferred to split these areas into independent models." A target selection factor set is obtained by creating a set through multiple selection factor rules. The target selection factor set is a set composed of multiple selection factor rules, which is used to guide the splitting scheme of the building model. Each selection factor rule represents a specific splitting strategy or priority. Based on the target selection factor set, a sub-building model splitting scheme corresponding to each selection factor rule is created. According to the requirements and priorities of each selection factor rule, the functional area, system composition and spatial structure are split and allocated to generate multiple sub-building model splitting schemes. Each splitting scheme is a sub-model generated according to a specific selection factor rule to meet specific design goals and requirements. According to the sub-building model splitting scheme corresponding to each selection factor rule, a first building model splitting scheme is generated. According to the splitting scheme of the sub-model, the target building model is divided into multiple sub-models, each of which has independent functional areas, system composition and spatial structure characteristics. For example, consider a BIM model of a commercial complex. By analyzing the functional area information, the server identifies the functional area features such as office area, commercial area and parking area. By extracting the system composition information, the server identifies the system composition features such as water supply system, air conditioning system and power system. By clustering the spatial structure information, the server identifies the spatial structure features such as floor division, room layout and spatial connection relationship.According to the selection factor rule matching, the server obtains multiple selection factor rules, such as "If the functional area is a commercial area, the system composition includes a water supply system and a power system, and the spatial structure feature is that multiple floors are divided into commercial areas and parking areas, then it is preferred to split these areas into independent models." Through set creation, the server forms a target selection factor set. Based on this set, the server creates a corresponding sub-building model splitting scheme for each selection factor rule. For example, according to the selection factor rules, the server assigns the commercial area and the parking area to independent sub-models to meet the design goals and requirements of the commercial complex. Finally, according to the splitting scheme of the sub-models, the server generated a splitting scheme for the first building model, dividing the BIM model of the commercial complex into multiple sub-models, each of which has independent functional areas, system composition and spatial structure characteristics.

[0107] S103, inputting the first building model splitting scheme and the target building BIM model into a preset model geometry separation network to perform model geometry splitting to obtain a plurality of first split BIM models;

[0108] It should be noted that the target building BIM model is subjected to geometric transformation to obtain target geometric data. The geometric transformation involves extracting geometric information from the BIM model and converting it into a format suitable for input into the model geometry separation network. The target geometric data and the first building model splitting scheme are input into a preset model geometry separation network. The model geometry separation network generally includes components such as a feature extraction layer, a separation layer, and an output layer, which are used to realize the splitting and separation of the geometric body. Through the feature extraction layer, the first building model splitting scheme is subjected to feature extraction to obtain feature information of multiple building model split virtual areas and each virtual area. These features may include geometric attributes such as the shape, size, and position of the area, as well as functions, system components, or other attribute information related to the area. Through the separation layer, the target geometric data is subjected to geometric separation according to the building model split virtual area and its feature information. The separation layer separates the target geometric data into multiple split geometries through geometric operations and algorithms based on the input virtual area information. Each split geometry represents an independent part or area in the first split BIM model. Through the output layer, a geometric label is created for each split geometry, and multiple split geometries are output as corresponding first split BIM models according to these labels. The geometric label can be used to identify the area, function, system composition and other information to which the split geometry belongs, so as to facilitate subsequent modeling, analysis or visualization. For example, consider a BIM model of a school building. The model is input into a preset model geometry separation network for splitting. A geometric conversion is performed to extract the geometric information in the school building BIM model and convert it into a format acceptable to the model geometry separation network. The converted target geometric data and the first building model splitting scheme are input into the model geometry separation network. Through the feature extraction layer, different functional areas in the school building model, such as classrooms, libraries and offices, are extracted, and feature information of these areas, such as size, shape and functional attributes, is obtained. Through the separation layer, the target geometric data is geometrically separated according to these feature information, and the school building model is separated into multiple split geometries, each of which represents an independent functional area. Through the output layer, geometry labels such as "classroom", "library" and "office" are created for each split geometry, and these split geometries are output as corresponding first split BIM models.

[0109] S104, performing model boundary extraction on the plurality of first split BIM models to obtain a plurality of model boundary features, and performing connection relationship analysis on the plurality of first split BIM models to obtain a plurality of connection relationship features;

[0110] Specifically, a plurality of first split BIM models are subjected to model boundary extraction to obtain a plurality of original boundary features. Model boundary extraction refers to extracting the boundary information of the model from the BIM model, which can be achieved by geometric calculation or boundary recognition algorithm. Each first split BIM model will obtain an original boundary feature, which describes the boundary shape and geometric properties of the model. At the same time, a connection relationship analysis is performed on the plurality of first split BIM models to obtain a plurality of original relationship features. Connection relationship analysis refers to analyzing the connection relationship between elements in the model, such as adjacent elements, nested relationships or hierarchical structures. By analyzing the connection relationship, the correlation and spatial relationship between different elements in the model can be identified. The plurality of original boundary features are input into a preset feature clustering model, and the cluster centers of these features are calculated using a preset first feature clustering function to obtain a first initial clustering center. Similarly, the plurality of original relationship features are input into a feature clustering model, and the cluster centers of these features are calculated using a preset second feature clustering function to obtain a second initial clustering center. The distance between each original boundary feature and the first initial clustering center is calculated to obtain a plurality of first clustering distances, and the mean operation is performed on these distances to obtain a first average distance. Similarly, the distance between each original relationship feature and the second initial clustering center is calculated to obtain multiple second clustering distances, and the mean operation is performed on these distances to obtain the second average distance. The parameters of the first feature clustering function are adjusted according to the first average distance to obtain the first target clustering function. Similarly, the parameters of the second feature clustering function are adjusted according to the second average distance to obtain the second target clustering function. The target clustering function adjusted in this way can better adapt to the clustering requirements of the original feature data. The first target clustering center is determined based on the first target clustering function, and the second target clustering center is determined based on the second target clustering function. These target clustering centers are calculated based on the adjusted target clustering function and represent the final center point of the feature clustering. The first target clustering center is used to perform secondary clustering on multiple original boundary features to obtain multiple model boundary features. These features describe the clustering results of model elements with similar boundary shapes and geometric attributes. Similarly, the second target clustering center is used to perform secondary clustering on multiple original relationship features to obtain multiple connection relationship features. These features describe the clustering results of model elements with similar connection relationships. For example, consider a BIM model of a large commercial center, which contains multiple first split BIM models. For each first split BIM model, perform model boundary extraction to obtain original boundary features, such as the boundary shape and area of ​​different shops. At the same time, perform connection relationship analysis to obtain original relationship features, such as the proximity relationship or hierarchical structure between shops. These original boundary features and original relationship features are input into the feature clustering model, and the cluster center is calculated using the preset clustering function. According to the distance between the cluster center and the original feature, the first cluster distance and the second cluster distance are calculated, and the mean distance is calculated.The parameters of the target clustering function are adjusted according to the mean distance to obtain the first target clustering function and the second target clustering function. Based on the target clustering function, the first target clustering center and the second target clustering center are determined. The original boundary features are clustered again using the first target clustering center to obtain the model boundary features, such as boundary clusters of different shapes and sizes of shops. At the same time, the original relationship features are clustered again using the second target clustering center to obtain the connection relationship features, such as the clustering of shops with similar proximity relationships or hierarchical structures.

[0111] S105, constructing a target boundary relationship matrix according to the plurality of model boundary features and the plurality of connection relationship features, and inputting the target boundary relationship matrix into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting solution;

[0112] Specifically, eigenvalue mapping is performed on multiple model boundary features and connection relationship features to obtain multiple boundary eigenvalues ​​and relationship eigenvalues. Eigenvalue mapping can be to convert the original features into eigenvalues ​​in numerical representation or other forms. This can make different features comparable and provide a basis for subsequent matrix conversion and analysis. Matrix conversion is performed on multiple boundary eigenvalues ​​and relationship eigenvalues ​​to obtain a target boundary relationship matrix. Matrix conversion is the process of organizing eigenvalues ​​into a matrix according to certain rules. For boundary features, they can be used as rows or columns of the matrix, corresponding to different model elements or boundary attributes. For relationship features, the relationship strength or similarity between model elements can be represented by the elements in the matrix. Through such matrix conversion, boundary features and relationship features can be combined to construct a target boundary relationship matrix. The target boundary relationship matrix is ​​input into a preset boundary interaction analysis model. The boundary interaction analysis model is generally composed of a first encoding network, a second encoding network and a decoding network. The encoding network is used to convert the input boundary relationship matrix into a low-dimensional representation to capture the features and relationships therein. The decoding network is used to reconstruct the boundary relationship matrix from the low-dimensional representation. Through this boundary interaction analysis model, the target boundary relationship matrix can be analyzed and predicted. Through the boundary interaction analysis model, the target boundary relationship matrix is ​​subjected to boundary interaction analysis to generate the target prediction probability. Boundary interaction analysis can identify the interactions and dependencies between boundaries and their impact on the splitting scheme. Through the learning and inference of the model, the interaction behaviors and possible splitting results between different boundaries can be predicted. According to the target prediction probability, the corresponding second building model splitting scheme is matched from multiple preset candidate building model splitting schemes. According to the size of the predicted probability, the splitting scheme that best meets the prediction can be selected as the splitting scheme of the second building model. In this way, the optimal splitting scheme can be selected to meet the design requirements and constraints based on the results of boundary relationship and interaction analysis. For example, consider a BIM model of a hospital building. Multiple model boundary features and connection relationship features are mapped into boundary eigenvalues ​​and relationship eigenvalues. These eigenvalues ​​are converted into a target boundary relationship matrix, where the rows and columns of the matrix correspond to different boundary features and model elements. The target boundary relationship matrix is ​​input into a preset boundary interaction analysis model, which can be a model based on deep learning or other algorithms. Through the boundary interaction analysis model, the target boundary relationship is analyzed to predict the interaction behaviors between different boundaries. According to the prediction result, a second building model splitting scheme that best meets the prediction can be selected from multiple candidate building model splitting schemes to meet the design requirements.

[0113] S106: Optimize model boundaries and connection relationships of the multiple first split BIM models according to the second building model splitting plan, generate multiple second split BIM models, and render and display the multiple second split BIM models.

[0114] Specifically, a plurality of first split BIM models are reorganized to generate a reorganized building BIM model. Model reorganization is a process of reorganizing a plurality of first split BIM models according to a second building model splitting scheme. By merging and adjusting the position, rotation and scaling of model elements, a reorganized building BIM model can be obtained, and the model includes the split parts defined by the second building model splitting scheme. According to the second building model splitting scheme, the model boundary parameters of the reorganized building BIM model are adjusted and the connection relationship is optimized to generate a plurality of second split BIM models. Model boundary parameter adjustment refers to fine-tuning the boundaries of the split parts to make them more in line with design requirements and spatial constraints. Connection relationship optimization refers to adjusting and optimizing the connection relationship between model elements to ensure coordination and consistency between the split parts. Through these adjustment and optimization operations, a plurality of second split BIM models that meet the requirements can be generated, each model representing a different splitting scheme. The plurality of second split BIM models are rendered to obtain a plurality of rendered BIM models. Rendering is the process of converting a building model into a visual image or graphic. By applying techniques such as lighting, materials, and textures, realistic rendering images can be generated for each second split BIM model. These rendered BIM models can show the appearance and spatial layout of different split schemes, providing a visual reference for design decisions and presentations. Multiple rendered BIM models are displayed in virtual reality. Virtual reality is a computer-generated simulation environment that allows users to experience the model in an immersive way through a head-mounted display, handle, or other interactive device. By combining the rendered BIM model with virtual reality technology, users can browse and interact freely in the virtual environment, gain in-depth insights into the details of different split schemes, and feel the atmosphere of the building space. For example, consider a BIM model of a residential project, which contains multiple first split BIM models, each of which represents an independent residential unit. These models are reorganized and merged into an overall residential project BIM model. According to the second building model split scheme, the boundaries of the reorganized models are adjusted and refined, the connection relationship is optimized, and multiple second split BIM models are generated, each of which represents a split residential unit. Each second split BIM model is rendered to generate realistic images that show the appearance and characteristics of each residential unit. Importing these rendered BIM models into a virtual reality platform allows designers and stakeholders to personally experience and explore the spatial layout and feel of each residential unit through VR devices.

[0115] In an embodiment of the present invention, the first building model splitting scheme and the target building BIM model are input into a preset model geometry separation network for model geometry splitting to obtain multiple first split BIM models; model boundary extraction is performed to obtain multiple model boundary features and connection relationship analysis is performed to obtain multiple connection relationship features; a target boundary relationship matrix is ​​constructed, and the target boundary relationship matrix is ​​input into a boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting scheme; model boundaries and connection relationships are optimized according to the second building model splitting scheme to generate multiple second split BIM models, and multiple second split BIM models are rendered and displayed. The present invention can generate a target selection factor set and a first building model splitting scheme by acquiring functional area information, system composition information and spatial structure information of the target building BIM model, and performing feature extraction and analysis. By inputting the first building model splitting scheme into a model geometry separation network for model geometry splitting, the method can generate multiple first split BIM models. Model boundary extraction and connection relationship analysis are performed on these models to further optimize the boundary and connection relationship of the model. By applying the boundary interaction analysis model and the target boundary relationship matrix, boundary interaction analysis can be performed and a second building model splitting scheme can be generated. By optimizing the model boundaries and connection relationships of multiple second split BIM models and applying rendering and display technology, the splitting results of the building can be understood and displayed more intuitively, thereby improving the splitting accuracy of the BIM model and the display accuracy of the split BIM model.

[0116] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0117] (1) Obtain the model configuration file of the target building BIM model, and extract the model configuration from the model configuration file to obtain the model configuration information to be matched;

[0118] (2) Performing functional feature matching on the configuration information of the model to be matched to obtain functional area information, and performing system composition matching on the configuration information of the model to be matched to obtain system composition information;

[0119] (3) Perform spatial distribution analysis on the target building BIM model to obtain the initial spatial distribution, and perform spatial classification on the target building BIM model to obtain the target spatial type;

[0120] (4) Determine the corresponding target space distribution according to the target space type, and perform collision detection on the initial space distribution and the target space distribution to obtain the collision detection result;

[0121] (5) Based on the collision detection results, a hierarchical structure analysis is performed on the target spatial distribution and the initial spatial distribution to generate the spatial structure information of the target building BIM model.

[0122] Specifically, the server obtains the model configuration file of the target building BIM model. The model configuration file is a metadata file of the building BIM model, which contains information such as the components, attributes and relationships of the building model. By obtaining the model configuration file, the structure and characteristics of the building model can be obtained. The model configuration file is subjected to model configuration extraction to obtain the model configuration information to be matched. Model configuration extraction refers to the process of extracting required information from the model configuration file. This can be achieved by parsing and analyzing the format and content of the model configuration file. The extracted information may include the type, size, material, attributes, etc. of the building elements. Functional feature matching is performed on the model configuration information to be matched to obtain functional area information. Functional feature matching is the process of comparing and matching the model configuration information with predefined functional features. By comparing the relationship between the configuration information of the building model and the known functional features, the functional attributes of each area can be determined. For example, by matching building elements with specific attributes (such as equipment, furniture, decoration, etc.), different functional areas such as offices, conference rooms, kitchens, etc. can be identified. System composition matching is performed on the model configuration information to be matched to obtain system composition information. System composition matching is the process of comparing and matching the model configuration information with predefined system components. By identifying and matching building elements with specific attributes, such as water supply systems, air conditioning systems, power systems, etc., the system composition information of the building model can be obtained. Perform spatial distribution analysis on the target building BIM model to obtain the initial spatial distribution. Spatial distribution analysis is to determine the position and layout of each area or element in the space by analyzing and calculating the geometric information of the building model. Through spatial distribution analysis, the relative position and spatial relationship of the elements of the building model can be obtained. Perform spatial classification on the target building BIM model to obtain the target space type. Spatial classification is the process of dividing the areas or elements in the model into different space types according to the functions and usage requirements of the building model. For example, the areas in the building model are divided into office areas, public areas, service areas, etc. Through spatial classification, corresponding planning and design strategies can be formulated for different space types. Determine the corresponding target space distribution according to the target space type, and perform collision detection on the initial space distribution and the target space distribution to obtain the collision detection result. According to the results of spatial classification, determine the layout requirements of the target space type and compare them with the initial space distribution. Through collision detection technology, analyze whether there is conflict or overlap between the elements in the building model. By detecting and solving collision problems, the spatial layout can be optimized to ensure the rationality and operability of the target spatial distribution. According to the collision detection results, the hierarchical structure analysis of the target spatial distribution and the initial spatial distribution is performed to generate the spatial structure information of the target building BIM model. By analyzing and sorting out the spatial distribution and collision detection results, the hierarchical structure of the building model can be established to determine the spatial relationships and organizational methods at different levels. This information can provide reference and basis for subsequent design and planning work.For example, consider a BIM model of an office building. Obtain the configuration file of the model and extract the model configuration information. Perform functional feature matching based on the model configuration information to identify different functional areas, such as offices, conference rooms, kitchens, etc. At the same time, perform system composition matching to determine system composition information such as water supply systems, air conditioning systems, and power systems. Perform spatial distribution analysis on the building model to determine the initial spatial distribution, such as the relative position and layout of different areas. Determine the target spatial distribution based on spatial classification and requirements, and perform collision detection to check whether there is a conflict. Generate spatial structure information of the target building BIM model through hierarchical structure analysis, such as the hierarchical relationship and organization of different floors and rooms.

[0123] In a specific embodiment, if Figure 2 As shown, the process of executing step S102 may specifically include the following steps:

[0124] S201, performing feature analysis on the functional area information to obtain a plurality of functional area features, wherein the plurality of functional area features include: area, function and attribute of the area;

[0125] S202, extracting features from the system composition information to obtain a plurality of system composition features, wherein the plurality of system composition features include: water supply system features, air conditioning system features, and power system features;

[0126] S203, performing feature clustering on the spatial structure information to obtain a plurality of spatial structure features, wherein the plurality of spatial structure features include: floor division, room layout, and spatial connection relationship;

[0127] S204, performing selection factor rule matching on multiple functional area features, multiple system composition features, and multiple spatial structure features to obtain multiple selection factor rules;

[0128] S205, creating a set through multiple selection factor rules to obtain a target selection factor set;

[0129] S206, creating a sub-building model splitting scheme corresponding to each selection factor rule through the target selection factor set;

[0130] S207: Generate a first building model splitting solution according to the sub-building model splitting solutions corresponding to each selection factor rule.

[0131] Specifically, the server performs feature analysis on the functional area information to obtain multiple functional area features. The functional area features may include descriptions of the area, function, and attributes. By analyzing and measuring each functional area, the area size of each area can be obtained. In addition, each area can be functionally classified according to functional requirements and usage planning, such as office area, conference room, rest area, etc. At the same time, the attributes of each area can also be identified, such as open, closed, privacy, etc. Feature extraction is performed on the system composition information to obtain multiple system composition features. The system composition features may include water supply system features, air conditioning system features, power system features, etc. By analyzing the elements and components in the building model, features related to these systems can be extracted. For example, water supply system features may include the location of water pipes, the distribution of valves, etc.; air conditioning system features may include the layout of air ducts, the location of vents, etc.; power system features may include the direction of cables, the arrangement of sockets, etc. Feature clustering is performed on the spatial structure information to obtain multiple spatial structure features. The spatial structure features may include descriptions of floor division, room layout, and spatial connection relationships. By analyzing the floor division and room layout in the building model, the distribution of different floors and the layout of the rooms can be obtained. At the same time, the spatial connection relationship between rooms, such as the location of doors and windows, the layout of passages, etc., can also be studied. The selection factor rules are matched for multiple functional area features, system composition features and spatial structure features to obtain multiple selection factor rules. The selection factor rule is a pre-defined rule set used to determine the corresponding splitting factors based on the matching results of different features. By comparing the matching degree between the functional area features, system composition features and spatial structure features and the selection factor rules, the applicable selection factor rules can be obtained. By creating a set through multiple selection factor rules, a target selection factor set can be obtained. The target selection factor set is based on the matching results of multiple selection factor rules, and determines the key factors in the building model splitting process. These factors can guide the creation and optimization of subsequent splitting schemes. According to the target selection factor set, a sub-building model splitting scheme corresponding to each selection factor rule is created. By interpreting and applying each selection factor rule, a corresponding splitting scheme can be created for each rule. These splitting schemes can guide the decomposition and organization of the building model to ensure that each area and system is properly split and configured. The first building model splitting scheme is generated according to the sub-building model splitting scheme corresponding to each selection factor rule. According to multiple splitting schemes, they are combined to obtain the splitting scheme of the first building model. This splitting scheme takes into account the influence of factors such as functional areas, system composition and spatial structure to ensure that the generated split model meets the expected design requirements. For example, considering a BIM model of an office building, the functional area information is first analyzed to obtain the functional area characteristics such as office area, conference room area, and rest area, including their area, function and attributes.Extract the features of system composition information, such as water supply system features, air conditioning system features, and power system features. Then, cluster the spatial structure information to obtain features such as floor division, room layout, and spatial connection relationship. According to the predefined selection factor rules, match the functional area features, system composition features, and spatial structure features to obtain a selection factor rule set. Through these selection factor rule sets, create a sub-building model splitting scheme corresponding to each rule. Based on these splitting schemes, generate a splitting scheme for the first building model to meet the requirements of architectural design and planning.

[0132] In a specific embodiment, if Figure 3 As shown, the process of executing step S103 may specifically include the following steps:

[0133] S301, performing geometric transformation on the target building BIM model to obtain target geometric data, and inputting the target geometric data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network includes: a feature extraction layer, a separation layer, and an output layer;

[0134] S302, extracting features from the first building model splitting solution through a feature extraction layer to obtain a plurality of building model splitting virtual areas and virtual area feature information corresponding to each building model splitting virtual area;

[0135] S303, performing geometric body separation on the target geometric body data according to the multiple building model split virtual areas and virtual area feature information corresponding to each building model split virtual area through the separation layer to obtain multiple split geometric bodies;

[0136] S304 . Create geometric tags for a plurality of split geometric bodies through the output layer, and output the plurality of split geometric bodies as a plurality of corresponding first split BIM models according to the geometric tags.

[0137] Specifically, the server performs geometric transformation on the target building BIM model and converts it into target geometric data. This includes extracting, processing and converting the geometric bodies in the building model to obtain geometric data representing various parts of the building. For example, the elements such as walls, floor slabs, beams and columns in the building model are converted into corresponding geometric representations. The target geometric data and the first building model splitting scheme are input into a preset model geometry separation network. The model geometry separation network is generally composed of a feature extraction layer, a separation layer and an output layer. Through the feature extraction layer, feature extraction is performed on the first building model splitting scheme. This may include feature extraction of each building model split virtual area in the splitting scheme to obtain virtual area feature information related to the area. For example, features such as shape, size, position, etc. of each area can be extracted and represented as feature vectors or feature descriptors. Through the separation layer, the target geometric data is geometrically separated according to the feature information of the building model split virtual area. The feature information is associated and separated from the target geometric data using machine learning or deep learning technology. For example, a neural network model can be used to take the input target geometry data and the corresponding virtual area feature information as input, and the target geometry data can be separated into multiple split geometries through network learning and inference. Through the output layer, geometry labels are created for the multiple split geometries and output as corresponding multiple first split BIM models. This can include assigning a unique identifier or label to each split geometry for subsequent processing and identification. For example, an independent ID or name can be assigned to each split geometry and saved together with other related attributes or metadata to build a complete first split BIM model. For example, consider a BIM model of a large commercial building, including multiple floors and different functional areas. The BIM model is geometrically transformed to convert elements such as walls, floor slabs, and columns into target geometry data. The target geometry data and the first building model splitting scheme are input into the model geometry separation network. Through the feature extraction layer, the features of each area in the splitting scheme, such as the shape, size, and position of the area, are extracted. Through the separation layer, the target geometry data is separated into multiple split geometries according to the regional features, and each split geometry represents an independent area. Through the output layer, a geometric label is created for each split geometry, for example, an independent ID is assigned to each area. Finally, through this process, multiple first split BIM models can be generated, each model corresponding to a split area, for subsequent processing and analysis.

[0138] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0139] (1) performing model boundary extraction on the plurality of first split BIM models to obtain a plurality of original boundary features, and performing connection relationship analysis on the plurality of first split BIM models to obtain a plurality of original relationship features;

[0140] (2) Inputting a plurality of original boundary features into a preset feature clustering model, and calculating cluster centers of the plurality of original boundary features by using a preset first feature clustering function to obtain a first initial cluster center, inputting a plurality of original relationship features into the feature clustering model, and calculating cluster centers of the plurality of original relationship features by using a preset second feature clustering function to obtain a second initial cluster center;

[0141] (3) calculating the distance between the multiple original boundary features and the first initial cluster center to obtain multiple first cluster distances and performing an average operation on the multiple first cluster distances to obtain a first average distance, and calculating the distance between the multiple original relationship features and the second initial cluster center to obtain multiple second cluster distances and performing an average operation on the multiple second cluster distances to obtain a second average distance;

[0142] (4) adjusting the parameters of the first feature clustering function according to the first average distance to obtain a first target clustering function, and adjusting the parameters of the second feature clustering function according to the second average distance to obtain a second target clustering function;

[0143] (5) determining a first target cluster center according to the first target cluster function, and determining a second target cluster center according to the second target cluster function;

[0144] (6) Performing secondary clustering on the multiple original boundary features through the first target clustering center to obtain multiple model boundary features, and performing secondary clustering on the multiple original relationship features through the second target clustering center to obtain multiple connection relationship features.

[0145] Specifically, model boundary extraction is performed on multiple first split BIM models to obtain multiple original boundary features. This includes performing boundary extraction operations on each split model to extract features such as the boundary shape, size and position of the model. For example, boundary information can be extracted from the geometric representation of the model through a boundary detection algorithm or a geometric calculation method. Connection relationship analysis is performed on multiple first split BIM models to obtain multiple original relationship features. This includes analyzing the connection relationship between the split models to identify the connection mode, interface location and mutual relationship between the models. For example, the connection relationship between the models can be inferred through methods such as geometric constraints, topological relationships or naming rules. Multiple original boundary features are input into a preset feature clustering model. This feature clustering model is usually composed of a feature extraction layer, a clustering layer and an output layer. Cluster centers are calculated for multiple original boundary features through a preset first feature clustering function to obtain a first initial clustering center. At the same time, multiple original relationship features are input into the feature clustering model, and cluster centers are calculated for multiple original relationship features through a preset second feature clustering function to obtain a second initial clustering center. Next, the distances between the multiple original boundary features and the first initial cluster center are calculated to obtain multiple first cluster distances, and the distances are averaged to obtain the first average distance. Similarly, the distances between the multiple original relationship features and the second initial cluster center are calculated to obtain multiple second cluster distances, and the distances are averaged to obtain the second average distance. The first feature clustering function is parameterized according to the first average distance to obtain the first target clustering function. Similarly, the second feature clustering function is parameterized according to the second average distance to obtain the second target clustering function. The first target clustering center is determined according to the first target clustering function, and the second target clustering center is determined according to the second target clustering function. These target clustering centers can be regarded as the center points representing the respective feature categories in the feature space. The multiple original boundary features are secondary clustered by the first target clustering center to obtain multiple model boundary features. The multiple original relationship features are secondary clustered by the second target clustering center to obtain multiple connection relationship features. These clustering operations can use clustering algorithms, such as K-means clustering or hierarchical clustering, to classify them into different clusters according to the similarity between the features. For example, suppose there is a building BIM model, which generates two first split BIM models after the first split: Model A and Model B. Model A is subjected to model boundary extraction to obtain original boundary features A1, A2, and A3. Model B is subjected to model boundary extraction to obtain original boundary features B1, B2, and B3. At the same time, a connection relationship analysis is performed on Model A and Model B to obtain original relationship features AB1, AB2, and AB3. The original boundary features A1, A2, and A3 are input into the feature clustering model, and the cluster center is calculated by the first feature clustering function to obtain the first initial cluster center A.Input the original relationship features AB1, AB2, and AB3 into the feature clustering model, calculate the cluster center through the second feature clustering function, and obtain the second initial cluster center AB. By calculating the distance between the original boundary features A1, A2, and A3 and the first initial cluster center A, the first cluster distance is obtained, and then the mean is calculated to obtain the first average distance. Similarly, the distance between the original relationship features AB1, AB2, and AB3 and the second initial cluster center AB is calculated to obtain the second cluster distance, and the mean is calculated to obtain the second average distance. According to the first average distance, the parameters of the first feature clustering function are adjusted to obtain the first target clustering function. According to the second average distance, the parameters of the second feature clustering function are adjusted to obtain the second target clustering function. The first target clustering function is used to determine the first target cluster center A, and the second target clustering function is used to determine the second target cluster center AB. The original boundary features A1, A2, and A3 are clustered again using the first target cluster center A to obtain the model boundary features A1', A2', and A3'. The original relationship features AB1, AB2, and AB3 are clustered again using the second target cluster center AB to obtain connection relationship features AB1', AB2', and AB3'. In this way, multiple optimized model boundary features and connection relationship features are obtained, which can be used for subsequent modeling and analysis.

[0146] In a specific embodiment, if Figure 4 As shown, the process of executing step S105 may specifically include the following steps:

[0147] S401, performing feature value mapping on multiple model boundary features and multiple connection relationship features to obtain multiple boundary feature values ​​and multiple relationship feature values;

[0148] S402, performing matrix conversion on multiple boundary eigenvalues ​​and multiple relationship eigenvalues ​​to obtain a target boundary relationship matrix;

[0149] S403, inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model includes: a first encoding network, a second encoding network and a decoding network;

[0150] S404, performing boundary interaction analysis on the target boundary relationship matrix through a boundary interaction analysis model to generate a target prediction probability;

[0151] S405 . Match a corresponding second building model splitting solution from a plurality of preset candidate building model splitting solutions according to the target prediction probability.

[0152] Specifically, the server performs eigenvalue mapping on multiple model boundary features and connection relationship features. Each eigenvalue is mapped to a suitable numerical range or standardized to ensure that they have consistent scale and comparability. Multiple boundary eigenvalues ​​and multiple relationship eigenvalues ​​are matrix-converted to generate a target boundary relationship matrix. In this step, the boundary eigenvalues ​​and relationship eigenvalues ​​can be organized into a matrix form in a certain order, wherein the rows of the matrix represent boundary features, the columns represent relationship features, and the matrix elements represent corresponding eigenvalues. The target boundary relationship matrix is ​​input into a preset boundary interaction analysis model, which is usually composed of a first encoding network, a second encoding network, and a decoding network. The first encoding network is used to learn the representation of boundary and relationship features, the second encoding network is used to extract the interaction information between features, and the decoding network is used to generate a target prediction probability. The boundary interaction analysis model is used to perform boundary interaction analysis on the target boundary relationship matrix to generate a target prediction probability. In this process, the model captures the correlation between boundary and relationship features by learning the interaction relationship between them, and generates a prediction probability to measure the matching degree of each boundary and relationship. According to the target prediction probability, the corresponding second building model splitting scheme is matched from multiple preset candidate building model splitting schemes. According to the prediction probability, the most matching building model splitting scheme can be selected as the final choice of the second building model. For example, suppose there is a building BIM model, which contains multiple model boundary features and connection relationship features. Model boundary features can include information such as boundary shape, size and position, while connection relationship features can include information such as interface position, connection mode and mutual relationship. For model boundary features, it is assumed that there are boundary features A, boundary features B and boundary features C. For connection relationship features, it is assumed that there are relationship features X, relationship features Y and relationship features Z. These features are mapped and normalized to ensure that they have consistent scale and comparability. The boundary feature values ​​and relationship feature values ​​are organized into a matrix form. It is assumed that the boundary feature values ​​are organized into a 3x3 matrix, where each row represents a boundary feature and each column represents a sample. At the same time, the relationship feature values ​​are organized into a 3x3 matrix, where each row represents a relationship feature and each column represents a sample. The two matrices are input into the preset boundary interaction analysis model. The model learns the interaction between boundary features and relationship features and generates a target prediction probability. Through the boundary interaction analysis model, the target prediction probability matrix is ​​obtained, in which each element represents the matching degree of the corresponding boundary and relationship. Assume that the obtained target prediction probability matrix is ​​a 3x3 matrix. According to the target prediction probability, the second building model splitting scheme that matches the scheme with the highest prediction probability is selected from the preset multiple candidate building model splitting schemes. For example, if the maximum value in the target prediction probability matrix appears in the second row and third column, then the corresponding candidate scheme is selected as the final choice of the second building model.

[0153] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0154] (1) reorganizing the multiple first split BIM models to obtain a reorganized building BIM model;

[0155] (2) According to the second building model splitting plan, adjust the model boundary parameters and optimize the connection relationship of the reorganized building BIM model to generate multiple second split BIM models;

[0156] (3) Rendering the plurality of second split BIM models to obtain a plurality of rendered BIM models, and performing virtual reality display on the plurality of rendered BIM models.

[0157] Specifically, the multiple first split BIM models are reorganized. This means that the components and elements in the multiple first split BIM models are recombined and integrated to generate a reorganized building BIM model. The reorganization method can be determined according to the design requirements and goals, such as readjusting the position of the components, adding new components, modifying the room layout, etc. According to the second building model splitting plan, the model boundary parameters of the reorganized building BIM model are adjusted and the connection relationship is optimized. According to the splitting plan, the boundary shape, size, height and other parameters of the building model can be adjusted, and the connection relationship between the components can be optimized to ensure the accuracy and rationality of the model. Through the model boundary parameter adjustment and connection relationship optimization, multiple second split BIM models can be generated, each model represents an optimized building splitting plan. These models can be used as candidate plans for subsequent evaluation and selection. Multiple second split BIM models are rendered to generate multiple rendered BIM models. Rendering is the process of converting a building model into a visual image or video to make it realistic and realistic. Through rendering, the appearance, material, lighting and other visual effects of the building can be presented to help designers and stakeholders better understand and evaluate the splitting plan. The rendered BIM model is displayed in virtual reality. Virtual reality is a simulation environment created by computer technology that allows users to interact and experience a virtual environment. By combining rendered BIM models with virtual reality technology, an immersive architectural experience can be achieved, allowing designers and stakeholders to freely move, observe, and evaluate the decomposition effect of the building in a virtual environment. For example, assume that there is a building BIM model that contains multiple first decomposition BIM models. According to design requirements, these decomposition models need to be reorganized and optimized according to the second building model decomposition plan. Model reorganization is performed on multiple first decomposition BIM models. Assume that the first decomposition model contains an independent floor, while the second decomposition model contains a smaller space unit. By recombining these two decomposition models, the independent floor can be embedded in the smaller space unit to form a new recombined building BIM model. Adjust and optimize the parameters of the recombined building BIM model according to the second building model decomposition plan. Assume that the decomposition plan requires adjusting the floor height and boundary shape of the recombined model. By adjusting the height of the floor and redefining the boundary, ensure that the model meets the requirements of the decomposition plan. Through model boundary parameter adjustment and connection relationship optimization, multiple second decomposition BIM models can be generated. For example, through different boundary adjustments and optimizations, a high-rise model and a low-rise model can be generated to meet different design needs and goals. Rendering and virtual reality display of multiple second split BIM models. Through rendering, these split models can be transformed into realistic images to show their appearance, materials and lighting effects.Through virtual reality technology, designers and stakeholders can personally experience and evaluate the effects of different spin-off plans in a virtual environment, such as roaming in the virtual environment, viewing details, and observing light and shadow effects.

[0158] The above describes the method for disassembling and displaying the BIM model in the embodiment of the present invention. The following describes the device for disassembling and displaying the BIM model in the embodiment of the present invention. Figure 5 In one embodiment of the present invention, a device for disassembling and displaying a BIM model includes:

[0159] The acquisition module 501 is used to acquire the functional area information and system composition information of the target building BIM model, and perform spatial structure analysis on the target building BIM model to obtain spatial structure information;

[0160] A creation module 502 is used to extract features from the functional area information, the system composition information and the spatial structure information to obtain a target selection factor set, and create a first building model splitting scheme according to the target selection factor set;

[0161] A splitting module 503 is used to input the first building model splitting scheme and the target building BIM model into a preset model geometry separation network to perform model geometry splitting to obtain a plurality of first split BIM models;

[0162] The processing module 504 is used to extract model boundaries of the multiple first split BIM models to obtain multiple model boundary features, and perform connection relationship analysis on the multiple first split BIM models to obtain multiple connection relationship features;

[0163] An analysis module 505 is used to construct a target boundary relationship matrix according to the multiple model boundary features and the multiple connection relationship features, and input the target boundary relationship matrix into a preset boundary interaction analysis model to perform boundary interaction analysis to generate a second building model splitting solution;

[0164] The display module 506 is used to optimize the model boundaries and connection relationships of the multiple first split BIM models according to the second building model splitting scheme, generate multiple second split BIM models, and render and display the multiple second split BIM models.

[0165] Optionally, the acquisition module 501 is specifically used for:

[0166] Obtain a model configuration file of the target building BIM model, and perform model configuration extraction on the model configuration file to obtain model configuration information to be matched;

[0167] Performing functional feature matching on the to-be-matched model configuration information to obtain functional area information, and performing system composition matching on the to-be-matched model configuration information to obtain system composition information;

[0168] Performing spatial distribution analysis on the target building BIM model to obtain an initial spatial distribution, and performing spatial classification on the target building BIM model to obtain a target spatial type;

[0169] Determining a corresponding target space distribution according to the target space type, and performing a collision detection on the initial space distribution and the target space distribution to obtain a collision detection result;

[0170] According to the collision detection result, a hierarchical structure analysis is performed on the target spatial distribution and the initial spatial distribution to generate spatial structure information of the target building BIM model.

[0171] Optionally, the creation module 502 is specifically used for:

[0172] Performing feature analysis on the functional area information to obtain a plurality of functional area features, wherein the plurality of functional area features include: area, function and attribute of the area;

[0173] Extracting features from the system composition information to obtain a plurality of system composition features, wherein the plurality of system composition features include: water supply system features, air conditioning system features, and power system features;

[0174] Performing feature clustering on the spatial structure information to obtain a plurality of spatial structure features, wherein the plurality of spatial structure features include: floor division, room layout, and spatial connection relationship;

[0175] Performing selection factor rule matching on the multiple functional area features, the multiple system composition features, and the multiple spatial structure features to obtain multiple selection factor rules;

[0176] Creating a set through the multiple selection factor rules to obtain a target selection factor set;

[0177] By using the target selection factor set, a sub-building model splitting scheme corresponding to each selection factor rule is created;

[0178] A first building model splitting scheme is generated according to the sub-building model splitting scheme corresponding to each selection factor rule.

[0179] Optionally, the splitting module 503 is specifically used for:

[0180] Performing geometric transformation on the target building BIM model to obtain target geometric data, and inputting the target geometric data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network includes: a feature extraction layer, a separation layer, and an output layer;

[0181] Performing feature extraction on the first building model splitting scheme through the feature extraction layer to obtain a plurality of building model splitting virtual areas and virtual area feature information corresponding to each building model splitting virtual area;

[0182] Through the separation layer, the target geometric body data is geometrically separated according to the multiple building model split virtual areas and virtual area feature information corresponding to each building model split virtual area to obtain multiple split geometric bodies;

[0183] Through the output layer, geometric tags of the multiple split geometric bodies are created, and the multiple split geometric bodies are output as corresponding multiple first split BIM models according to the geometric tags.

[0184] Optionally, the processing module 504 is specifically used for:

[0185] Performing model boundary extraction on the multiple first split BIM models to obtain multiple original boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple original relationship features;

[0186] Inputting the plurality of original boundary features into a preset feature clustering model, and calculating cluster centers of the plurality of original boundary features by using a preset first feature clustering function to obtain a first initial cluster center, inputting the plurality of original relationship features into the feature clustering model, and calculating cluster centers of the plurality of original relationship features by using a preset second feature clustering function to obtain a second initial cluster center;

[0187] Performing distance calculation on the multiple original boundary features and the first initial cluster center to obtain multiple first cluster distances and performing mean operation on the multiple first cluster distances to obtain a first average distance, and performing distance calculation on the multiple original relationship features and the second initial cluster center to obtain multiple second cluster distances and performing mean operation on the multiple second cluster distances to obtain a second average distance;

[0188] Adjusting the parameters of the first feature clustering function according to the first average distance to obtain a first target clustering function, and adjusting the parameters of the second feature clustering function according to the second average distance to obtain a second target clustering function;

[0189] Determining a first target cluster center according to the first target cluster function, and determining a second target cluster center according to the second target cluster function;

[0190] The plurality of original boundary features are clustered again using the first target cluster center to obtain a plurality of model boundary features, and the plurality of original relationship features are clustered again using the second target cluster center to obtain a plurality of connection relationship features.

[0191] Optionally, the analysis module 505 is specifically used for:

[0192] Performing feature value mapping on the multiple model boundary features and the multiple connection relationship features to obtain multiple boundary feature values ​​and multiple relationship feature values;

[0193] Performing matrix conversion on the plurality of boundary eigenvalues ​​and the plurality of relationship eigenvalues ​​to obtain a target boundary relationship matrix;

[0194] Inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model includes: a first encoding network, a second encoding network and a decoding network;

[0195] Performing boundary interaction analysis on the target boundary relationship matrix through the boundary interaction analysis model to generate a target prediction probability;

[0196] According to the target prediction probability, a corresponding second building model splitting scheme is matched from a plurality of preset candidate building model splitting schemes.

[0197] Optionally, the display module 506 is specifically used for:

[0198] Reorganizing the plurality of first split BIM models to obtain a reorganized building BIM model;

[0199] According to the second building model splitting scheme, adjusting model boundary parameters and optimizing connection relationships of the reorganized building BIM model to generate a plurality of second split BIM models;

[0200] The multiple second split BIM models are rendered to obtain multiple rendered BIM models, and the multiple rendered BIM models are displayed in virtual reality.

[0201] Through the collaborative cooperation of the above-mentioned components, the first building model splitting scheme and the target building BIM model are input into the preset model geometry separation network for model geometry splitting, and multiple first split BIM models are obtained; model boundary extraction is performed to obtain multiple model boundary features and connection relationship analysis is performed to obtain multiple connection relationship features; a target boundary relationship matrix is ​​constructed, and the target boundary relationship matrix is ​​input into the boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting scheme; model boundaries and connection relationships are optimized according to the second building model splitting scheme to generate multiple second split BIM models, and multiple second split BIM models are rendered and displayed. The present invention can generate a target selection factor set and a first building model splitting scheme by obtaining functional area information, system composition information and spatial structure information of the target building BIM model, and performing feature extraction and analysis. By inputting the first building model splitting scheme into the model geometry separation network for model geometry splitting, the method can generate multiple first split BIM models. Model boundary extraction and connection relationship analysis are performed on these models to further optimize the boundary and connection relationship of the model. Through the application of the boundary interaction analysis model and the target boundary relationship matrix, boundary interaction analysis can be performed and a second building model splitting scheme can be generated. By optimizing the model boundaries and connection relationships of multiple second split BIM models and applying rendering and display technology, the splitting results of the building can be understood and displayed more intuitively, thereby improving the splitting accuracy of the BIM model and the display accuracy of the split BIM model.

[0202] above Figure 5 The disassembly and display device of the BIM model in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The disassembly and display device of the BIM model in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0203] Figure 66 is a structural diagram of a BIM model disassembly and display device provided by an embodiment of the present invention. The BIM model disassembly and display device 600 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 610 (for example, one or more processors) and a memory 620, and one or more storage media 630 (for example, one or more mass storage devices) storing application programs 633 or data 632. Among them, the memory 620 and the storage medium 630 may be short-term storage or permanent storage. The program stored in the storage medium 630 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the BIM model disassembly and display device 600. Furthermore, the processor 610 may be configured to communicate with the storage medium 630 to execute a series of instruction operations in the storage medium 630 on the BIM model disassembly and display device 600.

[0204] The BIM model disassembly display device 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input and output interfaces 660, and / or one or more operating systems 631, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that Figure 6 The disassembled display equipment structure of the BIM model shown does not constitute a limitation on the disassembled display equipment of the BIM model, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0205] The present invention also provides a disassembled display device for a BIM model, the disassembled display device for a BIM model comprising a memory and a processor, the memory storing computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the disassembled display method for the BIM model in the above-mentioned embodiments.

[0206] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the method for disassembling and displaying the BIM model.

[0207] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0208] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0209] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for disassembling and displaying a BIM model, characterized in that: The split display method of the BIM model includes: Acquire functional area information and system composition information of the target building BIM model, and perform spatial structure analysis on the target building BIM model to obtain spatial structure information; Extracting features from the functional area information, the system composition information and the spatial structure information to obtain a target selection factor set, and creating a first building model splitting scheme according to the target selection factor set; specifically comprising: performing feature analysis on the functional area information to obtain multiple functional area features, wherein the multiple functional area features include: area, area function and area attribute; extracting features from the system composition information to obtain multiple system composition features, wherein the multiple system composition features include: water supply system features, air conditioning system features and power system features; performing feature clustering on the spatial structure information to obtain multiple spatial structure features, wherein the multiple spatial structure features include: floor division, room layout and spatial connection relationship; performing selection factor rule matching on the multiple functional area features, the multiple system composition features and the multiple spatial structure features to obtain multiple selection factor rules; creating a set through the multiple selection factor rules to obtain a target selection factor set; creating a sub-building model splitting scheme corresponding to each selection factor rule through the target selection factor set; generating a first building model splitting scheme according to the sub-building model splitting scheme corresponding to each selection factor rule; Inputting the first building model splitting scheme and the target building BIM model into a preset model geometry separation network for model geometry splitting to obtain a plurality of first split BIM models; specifically comprising: performing geometry conversion on the target building BIM model to obtain target geometry data, and inputting the target geometry data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network comprises: a feature extraction layer, a separation layer and an output layer; performing feature extraction on the first building model splitting scheme through the feature extraction layer to obtain a plurality of building model split virtual areas and virtual area feature information corresponding to each building model split virtual area; performing geometry separation on the target geometry data according to the plurality of building model split virtual areas and virtual area feature information corresponding to each building model split virtual area through the separation layer to obtain a plurality of split geometries; creating geometry labels for the plurality of split geometries through the output layer, and outputting the plurality of split geometries as corresponding plurality of first split BIM models according to the geometry labels; Performing model boundary extraction on the multiple first split BIM models to obtain multiple model boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple connection relationship features; A target boundary relationship matrix is ​​constructed according to the multiple model boundary features and the multiple connection relationship features, and the target boundary relationship matrix is ​​input into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting scheme; specifically comprising: performing eigenvalue mapping on the multiple model boundary features and the multiple connection relationship features to obtain multiple boundary eigenvalues ​​and multiple relationship eigenvalues; performing matrix conversion on the multiple boundary eigenvalues ​​and the multiple relationship eigenvalues ​​to obtain a target boundary relationship matrix; inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model comprises: a first encoding network, a second encoding network and a decoding network; performing boundary interaction analysis on the target boundary relationship matrix through the boundary interaction analysis model to generate a target prediction probability; matching the corresponding second building model splitting scheme from multiple preset candidate building model splitting schemes according to the target prediction probability; According to the second building model splitting scheme, the model boundaries and connection relationships of the multiple first split BIM models are optimized to generate multiple second split BIM models, and the multiple second split BIM models are rendered and displayed.

2. The method for disassembling and displaying a BIM model according to claim 1, characterized in that: The step of acquiring functional area information and system composition information of the target building BIM model and performing spatial structure analysis on the target building BIM model to obtain spatial structure information includes: Obtain a model configuration file of the target building BIM model, and perform model configuration extraction on the model configuration file to obtain model configuration information to be matched; Performing functional feature matching on the to-be-matched model configuration information to obtain functional area information, and performing system composition matching on the to-be-matched model configuration information to obtain system composition information; Performing spatial distribution analysis on the target building BIM model to obtain an initial spatial distribution, and performing spatial classification on the target building BIM model to obtain a target spatial type; Determining a corresponding target space distribution according to the target space type, and performing a collision detection on the initial space distribution and the target space distribution to obtain a collision detection result; According to the collision detection result, a hierarchical structure analysis is performed on the target spatial distribution and the initial spatial distribution to generate spatial structure information of the target building BIM model.

3. The method for disassembling and displaying a BIM model according to claim 1, characterized in that: The extracting model boundaries of the multiple first split BIM models to obtain multiple model boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple connection relationship features, include: Performing model boundary extraction on the multiple first split BIM models to obtain multiple original boundary features, and performing connection relationship analysis on the multiple first split BIM models to obtain multiple original relationship features; Inputting the plurality of original boundary features into a preset feature clustering model, and calculating cluster centers of the plurality of original boundary features by using a preset first feature clustering function to obtain a first initial cluster center, inputting the plurality of original relationship features into the feature clustering model, and calculating cluster centers of the plurality of original relationship features by using a preset second feature clustering function to obtain a second initial cluster center; Performing distance calculation on the multiple original boundary features and the first initial cluster center to obtain multiple first cluster distances and performing mean operation on the multiple first cluster distances to obtain a first average distance, and performing distance calculation on the multiple original relationship features and the second initial cluster center to obtain multiple second cluster distances and performing mean operation on the multiple second cluster distances to obtain a second average distance; Adjusting the parameters of the first feature clustering function according to the first average distance to obtain a first target clustering function, and adjusting the parameters of the second feature clustering function according to the second average distance to obtain a second target clustering function; Determining a first target cluster center according to the first target cluster function, and determining a second target cluster center according to the second target cluster function; The plurality of original boundary features are clustered again using the first target cluster center to obtain a plurality of model boundary features, and the plurality of original relationship features are clustered again using the second target cluster center to obtain a plurality of connection relationship features.

4. The method for disassembling and displaying a BIM model according to claim 1, characterized in that: The step of optimizing model boundaries and connection relationships of the plurality of first split BIM models according to the second building model splitting scheme, generating a plurality of second split BIM models, and rendering and displaying the plurality of second split BIM models includes: Reorganizing the plurality of first split BIM models to obtain a reorganized building BIM model; According to the second building model splitting scheme, adjusting model boundary parameters and optimizing connection relationships of the reorganized building BIM model to generate a plurality of second split BIM models; The multiple second split BIM models are rendered to obtain multiple rendered BIM models, and the multiple rendered BIM models are displayed in virtual reality.

5. A BIM model disassembly display device, characterized in that: The disassembled display device of the BIM model includes: An acquisition module is used to acquire functional area information and system composition information of a target building BIM model, and to perform spatial structure analysis on the target building BIM model to obtain spatial structure information; A creation module is used to extract features from the functional area information, the system composition information and the spatial structure information to obtain a target selection factor set, and create a first building model splitting scheme according to the target selection factor set; specifically comprising: performing feature analysis on the functional area information to obtain multiple functional area features, wherein the multiple functional area features include: area, area function and area attribute; performing feature extraction on the system composition information to obtain multiple system composition features, wherein the multiple system composition features include: water supply system features, air conditioning system features and power system features; performing feature clustering on the spatial structure information to obtain multiple spatial structure features, wherein the multiple spatial structure features include: floor division, room layout and spatial connection relationship; performing selection factor rule matching on the multiple functional area features, the multiple system composition features and the multiple spatial structure features to obtain multiple selection factor rules; performing set creation through the multiple selection factor rules to obtain a target selection factor set; creating a sub-building model splitting scheme corresponding to each selection factor rule through the target selection factor set; generating a first building model splitting scheme according to the sub-building model splitting scheme corresponding to each selection factor rule; A splitting module is used to input the first building model splitting scheme and the target building BIM model into a preset model geometry separation network for model geometry splitting to obtain a plurality of first split BIM models; specifically comprising: performing geometry conversion on the target building BIM model to obtain target geometry data, and inputting the target geometry data and the first building model splitting scheme into a preset model geometry separation network, wherein the model geometry separation network comprises: a feature extraction layer, a separation layer and an output layer; performing feature extraction on the first building model splitting scheme through the feature extraction layer to obtain a plurality of building model split virtual areas and virtual area feature information corresponding to each building model split virtual area; performing geometry separation on the target geometry data according to the plurality of building model split virtual areas and virtual area feature information corresponding to each building model split virtual area through the separation layer to obtain a plurality of split geometries; creating geometry labels for the plurality of split geometries through the output layer, and outputting the plurality of split geometries as corresponding plurality of first split BIM models according to the geometry labels; A processing module, configured to extract model boundaries from the plurality of first split BIM models to obtain a plurality of model boundary features, and perform connection relationship analysis on the plurality of first split BIM models to obtain a plurality of connection relationship features; The analysis module is used to construct a target boundary relationship matrix according to the multiple model boundary features and the multiple connection relationship features, and input the target boundary relationship matrix into a preset boundary interaction analysis model for boundary interaction analysis to generate a second building model splitting scheme; specifically comprising: performing eigenvalue mapping on the multiple model boundary features and the multiple connection relationship features to obtain multiple boundary eigenvalues ​​and multiple relationship eigenvalues; performing matrix conversion on the multiple boundary eigenvalues ​​and the multiple relationship eigenvalues ​​to obtain a target boundary relationship matrix; inputting the target boundary relationship matrix into a preset boundary interaction analysis model, wherein the boundary interaction analysis model comprises: a first encoding network, a second encoding network and a decoding network; performing boundary interaction analysis on the target boundary relationship matrix through the boundary interaction analysis model to generate a target prediction probability; matching the corresponding second building model splitting scheme from multiple preset candidate building model splitting schemes according to the target prediction probability; A display module is used to optimize the model boundaries and connection relationships of the multiple first split BIM models according to the second building model splitting scheme, generate multiple second split BIM models, and render and display the multiple second split BIM models.

6. A BIM model disassembly and display device, characterized in that: The BIM model disassembly and display device comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory so that the BIM model disassembly and display device executes the BIM model disassembly and display method according to any one of claims 1 to 4.

7. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the method for disassembling and displaying the BIM model as described in any one of claims 1 to 4 is implemented.

Citation Information

Patent Citations

  • BIM-based three-dimensional building model splitting processing method and device

    CN108427794A

  • BIM-based fabricated member parametrization design method

    CN108664697A

  • Building entity model lightweight method considering visual subject

    CN114998544A

  • Three-dimensional building model splitting processing method based on BIM

    CN115510545A

  • Three-dimensional model establishing method, device and system and medium

    CN116645488A