A lightweight processing method for railway BIM model

Abstract

The present invention discloses a railway BIM model lightweight processing method, which relates to the technical field of BIM model lightweight. The BIM model lightweight technology mentioned in the invention covers the lightweight conversion of the IFC standard BIM file format, and the subsequent steps of parameterized modeling, facet triangulation, similarity algorithm, and construction of a step-by-step loading construction mode that complies with the scene distance principle to further simplify the BIM model. Through the comprehensive application of these key points, the lightweight BIM model can improve performance, efficiency and operability while simplifying the complexity of the model, and meet the needs of specific application scenarios; through the BIM technology, the entire construction and maintenance process is dynamically simulated, and the complex construction and operation and maintenance processes can be intuitively displayed, so as to achieve scientific optimization of the construction and operation and maintenance plans, provide a real domestic digital base for the digital transformation of the railway four-electric engineering, and improve data integration, storage, governance, modeling, mining, analysis and circulation capabilities.

Description

Technical Field

[0001] The present invention relates to the technical field of lightweight BIM models, and in particular to a lightweight processing method for a railway BIM model. Background Art

[0002] With the improvement of the accuracy of the railway four-electric BIM model, large-scale BIM models such as 1G, 10G, or even dozens of G four-electric building models will have problems such as long model loading time and poor experience. In order to solve these problems, the present invention patent aims to achieve smooth loading of BIM 3D model files and lightweight BIM models, and the invention is described. We use two steps to process the lightweight of BIM models. The first is the lightweight processing of the standard BIM file format IFC format, which we convert into GLTF format; the second step is to use the WebGL API to perform more detailed lightweight processing on the BIM model while ensuring that the BIM model can meet the current scene display needs. First, we reasonably allocate data in the BIM model, including the allocation of geometric data and non-geometric data. The allocated geometric data uses parametric modeling to reduce details, simplifies the geometric model to remove unnecessary details and complexity, and divides the simplified model mesh into triangular facets. Then we use the Delaunay triangulation algorithm to convert these facets into smooth triangular meshes so that the facets meet the accuracy requirements. And adopt similarity algorithms to reduce the storage of model components, mainly by identifying and deleting redundant data to optimize model storage. By applying similarity algorithms, the model size can be reduced, thereby reducing storage and memory requirements. This method is usually applicable to parallel computing and distributed processing technologies for models with repeated structures and features to speed up model calculations. These methods are combined to improve model efficiency while maintaining model accuracy, providing reliable technical support for the lightweighting of railway four-electric BIM models.

[0003] The traditional railway four-electric BIM model has many problems such as large size, high cost and slow speed, and is prone to slow loading and high requirements for hardware equipment. Therefore, it is urgent to solve the problem of lightweight processing of BIM models. The BIM lightweight method we created adopts parametric modeling, compresses modeling data and accuracy, and adopts lightweight rendering algorithms to achieve lightweight processing of BIM models so that they can be displayed in browsers. The application of lightweight BIM models makes it vivid, intuitive and fast to obtain construction site data, and can grasp the quality, progress and investment completion of the project in real time without spending a lot of time and cost. This lightweight method provides more convenient and efficient technical support for project management and decision-making. Summary of the invention

[0004] The purpose of the present invention is to provide a railway BIM model lightweight processing method in order to solve the above-mentioned problems.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] According to one aspect of the present invention, a railway BIM model lightweight processing method is provided, comprising the following contents:

[0007] 1. Lightweight conversion of IFC BIM file format to GLTF format:

[0008] (1) The first step of file extraction uses the Xbim.Essentials component mentioned above. This component can parse IFC files and provide object-oriented access to IFC files. In other words, this component completes the extraction of IFC files. In subsequent operations, we can access IFC files just like accessing file objects in memory.

[0009] (2) The second step is file conversion. After the first step, the data in the memory is still organized in the IFC format. In the second step, the IFC format data in the memory must be extracted as needed so that the data can be loaded into the target format in the next step.

[0010] (3) The third step is to load the data extracted in the second step into the GLTF format. This step uses GLTF-CSHARP-LOADER to write the extracted data into the loader according to the GLTF organization format. The loader.save() function is called like a file writing function to write the data to the disk, or the GLTF format data is saved in the database to achieve fast transmission and loading of BIM models on the Internet.

[0011] 2. Perform subsequent lightweight processing on the BIM model after format conversion:

[0012] 2.1 Data allocation: BIM models are composed of 3D geometric data and non-geometric data such as model structure attributes. The process of separating these two parts of data can effectively remove about 20%-50% of the non-geometric data in the original BIM model file and output it as a data file for BIM application development;

[0013] 2.2 After the separation of geometric data and non-geometric data, the remaining 3D geometric data needs to be further lightweight and optimized to reduce the amount of data and the burden of rendering calculations on the client computer, thereby speeding up the download, rendering and function processing of the BIM model;

[0014] (a) Model parameterization: The parameterization process of the BIM model is to convert the various components in the building information model into a form that can be adjusted by defining parameters to achieve more efficient design and management. This process starts with data collection and modeling, covering the determination of parameterizable model features, such as wall thickness, window size, floor height, etc., and then using the functions of the BIM software to convert these features into adjustable parameters;

[0015] (b) Face triangulation: First, we decompose the complex geometry into faces, and then perform geometric simplification and optimization on the faces to remove redundant information and details. Next, we use the Delaunay triangulation algorithm to convert these faces into smooth triangular meshes to ensure that the model can achieve the required accuracy both visually and computationally. This processing not only reduces the storage space and data transmission requirements of the model, but also improves the efficiency and response speed of subsequent modeling, analysis and rendering processes;

[0016] (c) Similarity algorithm: In the railway four-electric BIM model, many components look exactly the same, but they are located in different positions or angles. In this case, the similarity algorithm can be used to merge the data, that is, only retain the data of one component, and other similar components only record a reference plus the spatial coordinates. This method can effectively reduce the storage volume of components and achieve the purpose of lightweighting;

[0017] (d) Construct a hierarchical component loading model that complies with the principle of scene distance: Create a multi-level component system and decompose the model into multiple levels: the distant level uses simplified geometry and low-resolution textures to quickly load and display the overall scene; while the near level contains more detailed and accurate component information, which is dynamically loaded according to the user's viewpoint to display high-quality details and textures, thereby providing a realistic visual experience while maintaining efficient performance;

[0018] 2.3. For the three-dimensional data processed by lightweight technology, the model is rendered at the micro level and macro level respectively;

[0019] At the micro level, we used a multi-level detail level, or LOD, technology. In the BIM lightweight process, this technology dynamically loads models with different levels of detail at runtime to optimize rendering performance.

[0020] Multi-level LOD, multiple levels of detail, by using different levels of geometry to represent objects, the model loaded at a distance is relatively simple, and the model loaded at a close distance is more refined;

[0021] At the macro level, the GPU parallel octree decomposition method is adopted. The octree is a tree data structure used for spatial data indexing. It recursively divides the three-dimensional space into eight sub-areas, each of which can be further divided. It divides a three-dimensional space area layer by layer into eight sub-cubes, each of which can be further divided into smaller sub-cubes until the required accuracy is achieved or specific conditions are met. Each node of the octree represents a cube area, and the node can have child nodes. The final leaf node represents the smallest area unit in the space. By combining the micro and macro levels, we have successfully optimized the lightweight BIM model and provided a practical and effective solution for the lightweight processing of the railway four-electric BIM model.

[0022] Furthermore, in 2.2, the parameterization of three-dimensional geometric data simplifies the complex geometric model in the railway four-electric system, removes unnecessary details and complexity, and retains only key structures and components. By simplifying the geometric model, the volume and complexity of the model can be greatly reduced, and the computing efficiency and model processing speed can be improved; the triangulation of the faces in the three-dimensional geometric data is to optimize the data structure of the model to improve the computing efficiency and performance while maintaining the accuracy of the geometric shape; the similarity algorithm processing in the geometric data can significantly reduce the complexity and data redundancy of the model, and improve the processing efficiency and performance of the model; and the step-by-step loading component mode that conforms to the principle of scene distance is adopted for processing.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. The BIM model lightweight technology mentioned in the invention covers the lightweight conversion of the IFC standard BIM file format, as well as the subsequent steps of parametric modeling, facet triangulation, similarity algorithm, and construction of a step-by-step loading construction mode that complies with the principle of scene distance to further simplify the BIM model. By comprehensively applying these key points, BIM model lightweighting can improve performance, efficiency and operability while simplifying the complexity of the model to meet the needs of specific application scenarios.

[0025] 2. Through the dynamic simulation of the entire construction and maintenance process through BIM technology, the complex construction and operation and maintenance processes can be intuitively displayed, and the scientific optimization of the construction and operation and maintenance plans can be achieved, providing a truly domestic digital foundation for the digital transformation of the railway four-electric engineering, and improving data integration, storage, governance, modeling, mining, analysis and circulation capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flowchart for lightweight processing of standard BIM files;

[0027] Figure 2Schematic diagram of the lightweight process of the model after format conversion. DETAILED DESCRIPTION

[0028] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0029] The patent of this invention is a technology for lightweight processing of BIM models of the four major subsystems of railway power supply, signal, communication and automation. It aims to simplify the complex railway system building information model, improve the efficiency of BIM model loading and display, and reduce the cost and time consumption of model processing.

[0030] according to Figure 1-Figure 2 As shown in the figure, this technical solution combines the characteristics of BIM (Building Information Model) and the railway four-electric system to achieve rapid modeling, analysis and optimization of the railway system. The following is a detailed description of the technical solution:

[0031] 1. Lightweight conversion of IFC BIM file format to GLTF format:

[0032] First of all, IFC is a storage specification for the construction industry, while GLTF is a storage specification for the computer field. GLTF only contains metadata related to 3D graphics. IFC not only contains 3D views of building models, but also contains a variety of attribute information, type information and other non-geometric data. When converting IFC files to GLTF files, in addition to removing non-geometric information, attention should also be paid to the data integrity of the geometric model in the initial IFC file. In converting IFC to lightweight GLTF format, we use the open source third-party development library xBIM. xBIM allows developers to read, create and view building information (BIM) models in IFC format. It fully supports geometry, topology operations and visualization.

[0033] Generally, we perform the following three steps: 1. Extract the file into memory; 2. Convert the file into the target format; 3. Load the data in memory into the target storage format and write it out. In order to convert the IFC file into a GLTF file, we also need to use Xbim.GLTF to do the above three steps. The specific plan is as follows:

[0034] The first step of file extraction uses the Xbim.Essentials component mentioned above, which can parse IFC files and provide object-oriented access to IFC files. In other words, this component completes the extraction of IFC files. In subsequent operations, we can access IFC files like accessing file objects in memory.

[0035] The second step is file conversion. After the first step, the data in the memory is still organized in IFC format. Therefore, in the second step, we need to extract the IFC format data in the memory as needed, so that we can load this data into the target format in the next step.

[0036] The third step is to load the data extracted in the second step into GLTF format. In this step, we will use GLTF-CSHARP-LOADER to do this. We only need to write the extracted data into the loader according to the GLTF organization form, and call the loader.save() function like calling the file writing function to write the data to the disk, or save the GLTF format data to the database, so as to achieve fast transmission and loading of BIM models on the Internet.

[0037] The following is a size comparison of the file loaded into GLTF and the original IFC file after geometric data extraction. It can be seen that the model file can be compressed to about 50% after the first step.

[0038]

[0039] 2. Perform subsequent lightweight processing on the BIM model after format conversion

[0040] (1) Data allocation: BIM models are composed of three-dimensional geometric data and non-geometric data such as model structure attributes. The process of separating these two parts of data can effectively remove about 20%-50% of the non-geometric data in the original BIM model file and output it as a data file for use in BIM application development.

[0041] (2) After the separation of geometric data and non-geometric data, the remaining three-dimensional geometric data needs to be further lightweight and optimized to reduce the amount of data and the burden of rendering calculations on the client computer, thereby speeding up the download, rendering and functional processing of the BIM model. The parameterization of three-dimensional geometric data is to simplify the complex geometric model in the railway four-electric system, remove unnecessary details and complexity, and retain only the key structures and components. By simplifying the geometric model, the volume and complexity of the model can be greatly reduced, and the calculation efficiency and model processing speed can be improved; the triangulation of the face in the three-dimensional geometric data is to optimize the data structure of the model to improve the calculation efficiency and performance while maintaining the accuracy of the geometric shape; the similarity algorithm processing in the geometric data can significantly reduce the complexity and data redundancy of the model, and improve the processing efficiency and performance of the model; and in order to load the built BIM model more smoothly, so that the key attributes of the model can be seen from the overall and detailed aspects, we adopt the step-by-step loading component mode that conforms to the principle of scene distance to handle it.

[0042] (a) Model parameterization: The parameterization process of BIM models is to convert the various components in the building information model into a form that can be adjusted by defining parameters to achieve more efficient design and management. This process starts with data collection and modeling, covering the identification of parameterizable model features, such as wall thickness, window size, floor height, etc., and then using the functions of BIM software to convert these features into adjustable parameters. By adjusting these parameters, designers can quickly optimize and customize the model to meet the needs of different design solutions while maintaining the consistency and accuracy of the model. This approach not only improves the flexibility and responsiveness of the design process, but also makes the model more efficient and controllable in subsequent analysis, collaboration and construction stages.

[0043] (b) Face triangulation: First, we decompose the complex geometry into faces, and then perform geometric simplification and optimization on the faces to remove redundant information and details. Next, we use the Delaunay triangulation algorithm to convert these faces into smooth triangular meshes to ensure that the model can achieve the required accuracy both visually and computationally. This processing not only reduces the storage space and data transmission requirements of the model, but also improves the efficiency and response speed of subsequent modeling, analysis, and rendering.

[0044] Among them, the Delaunay triangulation algorithm is a classic geometric algorithm used to convert a given set of points (or patches) into a triangular mesh to ensure the geometric stability and efficiency of the mesh. It first imports the point set, that is, first takes the vertex set or patch in the BIM model as input, and converts it into a set of points in the plane or space as needed. Then the algorithm calculates a set of non-overlapping triangles based on a set of patches to cover the entire geometric shape, ensuring that among all the generated triangles, no triangle's circumscribed circle contains any other vertices. This property makes the generated triangular mesh more stable and geometrically optimized. Finally, some optimization steps are usually performed, such as removing unnecessary triangles or optimizing boundary conditions, to ensure that the generated mesh reaches the required accuracy both visually and computationally.

[0045] (c) Similarity algorithm: In the railway BIM model, many components look exactly the same, but they are located in different positions or angles. In this case, the similarity algorithm can be used to merge the data, that is, only retain the data of one component, and record only a reference plus spatial coordinates for other similar components. For example, if we retain the data of one cylinder, we only need to record a reference + spatial coordinates for other cylinders. This method can effectively reduce the storage volume of components and achieve the purpose of lightweighting.

[0046] (d) Construct a step-by-step component loading mode that complies with the principle of scene distance: For BIM models with large data volumes, which contain a large number of components, it is unrealistic to completely download and load them in a web browser. At the same time, the field of view of the BIM model is limited. Therefore, this feature can be used to create a multi-level component system by decomposing the model into multiple levels: the distant level uses simplified geometry and low-resolution textures to quickly load and display the overall scene; while the near-view level contains more detailed and accurate component information, which is dynamically loaded according to the user's viewpoint to display high-quality details and textures, thereby providing a realistic visual experience while maintaining efficient performance. In this way, when users browse the BIM model, they can see the overall scene from a distance without detailed details, and can see specific details from a close distance without having to load the entire model. This method significantly improves the loading speed and user experience of BIM models in web browsers, and solves the problem of lightweighting BIM models with large data volumes.

[0047] (3) For the three-dimensional data processed by lightweight technology, we then render it in real time on the client. Rendering can present the simplified BIM model in a high-quality visual way, enhancing its visual appeal and expressiveness. Here we render the model at the micro level and macro level respectively. At the micro level, we use a multi-level detail level (LOD) technology. During the BIM lightweight process, models with different levels of detail are dynamically loaded at runtime to achieve the purpose of optimizing rendering performance. Multi-level LOD (multi-level detail) uses different levels of geometry to represent objects. Models loaded at a long distance are relatively simple, while models loaded at a close distance are more refined. This method improves display efficiency and reduces storage requirements without compromising visual effects. In a single rendering, the volume (i.e., the amount of rendering work) can be expressed as the number of primitives multiplied by the primitive precision. When the viewpoint is far away, although the number of primitives is large, the primitive precision is low, so the overall volume is still controllable. On the contrary, when the viewpoint distance is close, although the accuracy of the primitives is high, the actual number of primitives is small, so the volume can still be kept within a controllable range. At the macro level, how can we quickly select the components needed in the scene from the huge number of components for rendering? We adopted the GPU parallel octree decomposition method. The octree is a tree data structure for spatial data indexing. It recursively divides the three-dimensional space into eight sub-areas, each of which can be further divided. It divides a three-dimensional space area layer by layer into eight sub-cubes, each of which can continue to be divided into smaller sub-cubes until the required accuracy is achieved or specific conditions are met. Each node of the octree represents a cube area, the node can have child nodes, and the final leaf node represents the smallest area unit in the space. By combining the micro and macro levels, we successfully optimized the lightweight BIM model and provided a practical and effective solution for the lightweight processing of the railway four-electric BIM model. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and range of equivalent elements of the claims are included in the present invention. Any reference sign in the claims should not be regarded as limiting the claims involved.

[0048] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A railway BIM model lightweight processing method, characterized by: Includes the following:

1. Lightweight conversion of IFC BIM file format to GLTF format: (1) The first step of file extraction uses the Xbim.Essentials component mentioned above, which parses the IFC file and provides an object-oriented access method to access the IFC file; (2) The second step is file conversion. After the first step, the data in the memory is still organized in the IFC format. In the second step, the IFC format data in the memory must be extracted as needed so that the data can be loaded into the target format in the next step. (3) The third step is to load the data extracted in the second step into the GLTF format, and use GLTF-CSHARP-LOADER to process the extracted data and write it into the loader according to the GLTF organization form. The loader.save() function can be called like a file writing function to write the data to the disk, or the GLTF format data can be saved in the database to achieve fast transmission and loading of BIM models on the Internet; 2. Perform subsequent lightweight processing on the BIM model after format conversion: 2.1 Data allocation: BIM models are composed of 3D geometric data and non-geometric data such as model structure attributes. The process of separating these two parts of data can effectively remove about 20%-50% of the non-geometric data in the original BIM model file and output it as a data file for BIM application development; 2.2 After the separation of geometric data and non-geometric data, the remaining 3D geometric data needs to be further lightweight and optimized to reduce the amount of data and reduce the burden of rendering calculations on the client computer; (a) Model parameterization: The parameterization process of the BIM model is to convert the various components in the building information model into a form that can be adjusted by defining parameters, so as to achieve more efficient design and management; (b) Face triangulation: First, complex geometry is decomposed into faces, and then the faces are geometrically simplified and optimized to remove redundant information and details. The faces are then converted into smooth triangular meshes using the Delaunay triangulation algorithm to ensure that the model can achieve the required accuracy both visually and computationally. (c) Similarity algorithm: Similarity algorithm is used to merge data of the same components, effectively reducing the storage volume of components and achieving the purpose of lightweighting; (d) Construct a hierarchical component loading model that complies with the principle of scene distance: Create a multi-level component system and decompose the model into multiple levels: the distant level uses simplified geometry and low-resolution textures to quickly load and display the overall scene; while the near level contains more detailed and accurate component information, which is dynamically loaded according to the user's viewpoint to display high-quality details and textures; 2.3 For the three-dimensional data processed by lightweight technology, the model is rendered at the micro level and macro level respectively; At the micro level, using the Levels of Detail LOD technology, by using different levels of geometry to represent objects, the models loaded at a long distance are relatively simplified, while the models loaded at a close distance are more refined; At the macro level, GPU parallel octree decomposition is adopted: initially, the entire space is represented as a cube, which is divided into eight sub-cubes; each sub-cube is further divided into eight sub-cubes as needed, and this process is performed recursively until the predetermined maximum depth is reached or the number of points in each cube is lower than the threshold; each node of the octree represents a cube area, the node has child nodes, and the final leaf node represents the smallest area unit in the space. The lightweight BIM model is optimized by combining the micro and macro levels, providing a practical and effective solution for the lightweight processing of the railway four-electric BIM model.

2. A railway BIM model lightweight processing method as claimed in claim 1, characterized in that: In 2.2, the further optimization of three-dimensional geometric data is to simplify the complex geometric model in the railway four-electric system, remove unnecessary details and complexity, and retain only key structures and components. By simplifying the geometric model, the volume and complexity of the model are greatly reduced, and the computing efficiency and model processing speed are improved; the triangulation of the patches in the three-dimensional geometric data is to optimize the data structure of the model on the basis of maintaining the accuracy of the geometric shape to improve the computing efficiency and performance; the similarity algorithm processing in the geometric data can significantly reduce the complexity and data redundancy of the model, and improve the processing efficiency and performance of the model; and the step-by-step loading component mode that conforms to the principle of scene distance is adopted for processing.

Images