Engineering BIM model splitting and recombining method and system based on AI

Abstract

The invention provides an engineering BIM model splitting and recombining method and system based on AI, and relates to the technical field of BIM, BIM model data is cleaned, attribute information of components is extracted from a BIM model, geometric information in the BIM model is extracted by using a computer vision technology, and component feature data is jointly formed; the method comprises the following steps: analyzing a construction task in a WBS file by using a natural semantic processing technology to form task feature data, completing matching of the task feature data and model component feature data by using an AI model to obtain a mapping table of tasks and components, and obtaining corresponding components according to the mapping table of the tasks and the components and the construction task. According to the method, the BIM model is divided into the sub-models corresponding to the components, the corresponding sub-models are dynamically recombined on the basis of requirements, the corresponding combined model is generated, the combined model is presented through the three-dimensional visual interface, the problem that a large amount of manpower needs to be consumed in existing BIM model splitting and recombining is solved, and the method is suitable for BIM model splitting and recombining.

Description

Technical Field

[0001] The present invention relates to the field of BIM technology, and particularly to a method and system for splitting and reorganizing engineering BIM models based on AI. Background Art

[0002] Through creating a three-dimensional model containing the physical characteristics and functional attributes of a project, BIM technology integrates the data information in each stage of the project and has been widely applied in engineering design and construction. However, in the engineering design stage, the BIM model mainly focuses on the expression and verification of design intentions, pays attention to the accuracy of geometric shapes and the rationality of design parameters. The model organization is often based on specialties, such as models are respectively established for specialties like architecture, structure, mechanical and electrical, etc., and are integrated through coordination tools. In the engineering construction stage, the BIM model is mainly used for guiding construction, schedule control, cost management, etc., and needs to support specific construction activities. The model organization is more based on construction processes and functional modules. When the engineering BIM model is deeply applied in the construction management process, it is necessary to perform necessary splitting or merging processing on model elements according to the work breakdown structure (WBS) of the construction work and the construction method. However, the traditional BIM model management method usually relies on manual operations for model splitting and reorganization, with complex operations, low efficiency, easy to make mistakes, and unable to adapt to the changes in the construction progress in real time.

[0003] CN116628826A discloses a method and system for quickly splitting a BIM model. It mainly obtains the three-dimensional information and attribute information of the BIM model, selects split components (vertical components or horizontal components), and splits the model components into multiple model sub-components according to different splitting methods (splitting by reference lines, splitting by set spacing, splitting by fixed spacing, splitting by fixed number of segments), performs merging according to actual needs, and can perform operations such as renaming and viewing the engineering quantity statement on the split and merged models, and stores them in the form of automatically uploading the updated BIM model.

[0004] The BIM model splitting and reorganization scheme mainly relies on static rules or manual operations, with a cumbersome process, low efficiency and easy to make mistakes. In addition, since construction is a dynamic process, the traditional model management method is difficult to adapt to the changes in the construction progress in real time, the model update progress lags behind, and the existing technical solutions lack intelligent decision-making support and are difficult to cope with the complex and changeable construction requirements in large-scale projects. Summary of the Invention

[0005] The technical problem to be solved by the present invention: The present invention provides a method and system for splitting and reorganizing engineering BIM models based on AI, which solves the problem that a large amount of manpower is required for the existing BIM model splitting and reorganization.

[0006] The technical solution adopted by the present invention to solve the above technical problems: An AI-based engineering BIM model splitting and recombination method, comprising the following steps:

[0007] S1. Clean the BIM model data, extract the attribute information of components from the BIM model and extract the geometric information in the BIM model using computer vision technology, jointly forming component feature data;

[0008] S2. Use natural semantic processing technology to parse the construction tasks in the WBS file, extract construction content, construction area, materials and equipment, forming task feature data;

[0009] S3. Use the AI model to complete the matching of task feature data and model component feature data, obtaining a mapping table of tasks and components;

[0010] S4. According to the mapping table of tasks and components, obtain the corresponding components according to the construction tasks, and split the BIM model into sub-models corresponding to the components;

[0011] S5. Based on the requirements of the construction stage or different task work packages, dynamically recombine the corresponding sub-models to generate corresponding combined models;

[0012] S6. Present the combined model through a three-dimensional visualization interface.

[0013] Further, S6 further includes color-coding the sub-models in the combined model according to the construction status.

[0014] The present invention also provides an engineering BIM model splitting and recombination system based on an AI model to implement the above-mentioned engineering BIM model splitting and recombination method based on an AI model. The system includes a BIM model parsing module, a WBS task parsing module, a task-component mapping module, an automatic BIM model splitting module, a sub-model recombination module, and a visualization interaction module;

[0015] The BIM model parsing module cleans the BIM model data, extracts the attribute information of components from the BIM model and extracts the geometric information in the BIM model using computer vision technology, jointly forming model component feature data;

[0016] The WBS task parsing module uses natural semantic processing technology to parse the construction tasks in the WBS file, extracts construction content, construction area, materials and equipment, forming task feature data;

[0017] The task-component mapping module uses the AI model to complete the matching of task feature data and model component feature data, obtaining a mapping table of tasks and components;

[0018] The BIM model automatic splitting module obtains corresponding components according to the construction tasks based on the mapping table between tasks and components, and splits the BIM model into sub-models corresponding to the components;

[0019] The sub-model recombination module dynamically recombines the corresponding sub-models based on the requirements of construction stages or different task work packages to generate corresponding combined models;

[0020] The visualization interaction module presents the combined model through a three-dimensional visualization interface.

[0021] Furthermore, the visualization interaction module is also used to color-label the sub-models in the combined model according to the construction status.

[0022] Furthermore, the geometric information includes shape, size, and position.

[0023] Furthermore, the attribute information includes category, material, and construction stage.

[0024] Furthermore, the AI model is trained using the supervised learning method to learn the matching relationship between historical task feature data and component feature data.

[0025] The beneficial effects of the present invention: The present invention provides an AI-based engineering BIM model splitting and recombination method and system. By cleaning the BIM model data, extracting the attribute information of components from the BIM model and using computer vision technology to extract the geometric information in the BIM model to jointly form component feature data, using natural semantic processing technology to parse the construction tasks in the WBS file, extracting construction content, construction area, materials, and equipment to form task feature data, using the AI model to complete the matching of task feature data and model component feature data, obtaining the mapping table between tasks and components, obtaining the corresponding components according to the construction tasks based on the mapping table between tasks and components, and splitting the BIM model into sub-models corresponding to the components, dynamically recombining the corresponding sub-models based on the requirements of construction stages or different task work packages to generate corresponding combined models, and presenting the combined models through a three-dimensional visualization interface, which solves the problem that the existing BIM model splitting and recombination requires a large amount of manpower. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a schematic flowchart of an AI-based engineering BIM model splitting and recombination method provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] In view of the problem that the existing BIM model splitting and recombination requires a large amount of manpower, the present invention provides an AI-based engineering BIM model splitting and recombination method, as Figure 1 shown, including the following steps:

[0028] S1. Clean the BIM model data, extract the attribute information of components from the BIM model, and use computer vision technology to extract the geometric information in the BIM model to jointly form component feature data. Specifically, the geometric information includes shape, size, and position, and the attribute information includes category, material, and construction stage.

[0029] S2. Use natural language processing technology to parse the construction tasks in the WBS file, extract construction content, construction area, materials, and equipment to form task feature data.

[0030] S3. Use the AI model to complete the matching of task feature data and model component feature data to obtain a mapping table of tasks and components. Specifically, the AI model is trained using the supervised learning method to learn the matching relationship between historical task feature data and component feature data, and uses the matching task feature data and component feature data as the data set to train the AI model.

[0031] S4. According to the mapping table of tasks and components, obtain the corresponding components according to the construction tasks, and split the BIM model into sub-models corresponding to the components.

[0032] S5. Based on the requirements of the construction stage or different task work packages, dynamically reorganize the corresponding sub-models to generate corresponding combined models.

[0033] S6. Present the combined model through a three-dimensional visualization interface. Since each sub-model in the combined model corresponds to a construction task, the sub-models corresponding to the completed, ongoing, and unstarted tasks can be color-coded according to the on-site construction status to intuitively display the construction progress.

[0034] The present invention also provides an AI-based engineering BIM model splitting and reorganization system to implement the above-mentioned AI model-based engineering BIM model splitting and reorganization method. The system includes a BIM model parsing module, a WBS task parsing module, a task-component mapping module, an automatic BIM model splitting module, a sub-model reorganization module, and a visualization interaction module.

[0035] The BIM model parsing module cleans the BIM model data, extracts the attribute information of components from the BIM model, and uses computer vision technology to extract the geometric information in the BIM model to jointly form model component feature data.

[0036] The WBS task parsing module uses natural language processing technology to parse the construction tasks in the WBS file, extracts construction content, construction area, materials, and equipment to form task feature data.

[0037] The task and component mapping module uses an AI model to complete the matching of task feature data and model component feature data, and obtains a mapping table of tasks and components;

[0038] The BIM model automatic splitting module obtains the corresponding components according to the construction tasks based on the mapping table of tasks and components, and splits the BIM model into sub-models corresponding to the components;

[0039] The sub-model recombination module dynamically recombines the corresponding sub-models based on the requirements of the construction stage or different task work packages to generate corresponding combined models;

[0040] The visualization interaction module presents the combined model through a three-dimensional visualization interface.

[0041] Specifically, the geometric information includes shape, size, and position; the attribute information includes category, material, and construction stage; the AI model is trained using the supervised learning method to learn the matching relationship between historical task feature data and component feature data, that is, using the matched task feature data and component feature data as the data set to train the AI model. Each sub-model in the combined model corresponds to a construction task, and the sub-models corresponding to the completed, ongoing, and unstarted tasks can be color-coded according to the on-site construction status to intuitively display the construction progress.

Claims

1. The method for splitting and reorganizing engineering BIM models based on AI models is characterized by: The following steps are involved: S1. Clean the BIM model data, extract the attribute information of the components from the BIM model, and use computer vision technology to extract the geometric information in the BIM model to form component feature data; S2. Use natural semantic processing technology to parse the construction tasks in the WBS file, extract the construction content, construction area, materials and equipment, and form task feature data; S3. Use the AI ​​model to match the task feature data with the model component feature data to obtain a mapping table between the task and the component; S4. According to the mapping table between tasks and components, corresponding components are obtained according to the construction tasks, and the BIM model is split into sub-models corresponding to the components; S5. Based on the requirements of the construction phase or different task work packages, the corresponding sub-models are dynamically reorganized to generate the corresponding combined model; S6. Present the combined model through a three-dimensional visualization interface.

2. The method for splitting and reorganizing an engineering BIM model based on an AI model according to claim 1, characterized in that: S6 also includes color marking of sub-models in the combined model according to the construction status.

3. The method for splitting and reorganizing an engineering BIM model based on an AI model according to claim 1, characterized in that: The geometric information includes shape, size and position.

4. The method for splitting and reorganizing an engineering BIM model based on an AI model according to claim 1, characterized in that: The attribute information includes category, material and construction stage.

5. The method for splitting and reorganizing an engineering BIM model based on an AI model according to claim 1, characterized in that: The AI ​​model is trained using supervised learning methods to learn the matching relationship between historical task feature data and component feature data.

6. The engineering BIM model splitting and reorganization system based on AI model is characterized by: Implementing the engineering BIM model splitting and reorganization method based on the AI ​​model as claimed in claim 1, the system includes a BIM model parsing module, a WBS task parsing module, a task and component mapping module, a BIM model automatic splitting module, a sub-model reorganization module and a visual interaction module; The BIM model parsing module cleans the BIM model data, extracts the attribute information of the components from the BIM model, and uses computer vision technology to extract the geometric information in the BIM model, thereby forming model component feature data. The WBS task parsing module uses natural semantic processing technology to parse the construction tasks in the WBS file, extract the construction content, construction area, materials and equipment, and form task feature data; The task and component mapping module uses the AI ​​model to complete the matching of task feature data and model component feature data to obtain a mapping table of tasks and components; The BIM model automatic splitting module obtains corresponding components according to the construction tasks based on the mapping table between tasks and components, and splits the BIM model into sub-models corresponding to the components; The sub-model reorganization module dynamically reorganizes the corresponding sub-models based on the requirements of the construction stage or different task work packages to generate a corresponding combined model; The visualization interaction module presents the combined model through a three-dimensional visualization interface.

7. The engineering BIM model splitting and reorganization system based on the AI ​​model according to claim 6 is characterized in that: The visualization interaction module is also used to color-code the sub-models in the combined model according to the construction status.

8. The engineering BIM model splitting and reorganization system based on the AI ​​model according to claim 6 is characterized in that: The geometric information includes shape, size and position.

9. The engineering BIM model splitting and reorganization system based on the AI ​​model according to claim 6 is characterized in that: The attribute information includes category, material and construction stage.

10. The engineering BIM model splitting and reorganization system based on the AI ​​model according to claim 6 is characterized in that: The AI ​​model is trained using supervised learning methods to learn the matching relationship between historical task feature data and component feature data.

Images