A Multi-Level Semantic Constraint-Based BIM-GIS Virtual-Real Mapping Method and System for High-Speed ​​Railway Infrastructure

Abstract

This invention discloses a multi-level semantic constraint BIM-GIS virtual-real mapping method and system for high-speed railway infrastructure, belonging to the field of railway intelligent technology. Based on the acquired high-speed railway infrastructure BIM model, this invention establishes an IFC component library, an information library, and the mapping relationship between them. The IFC component library is then filtered, removed, and updated based on filtering rules to obtain a modified IFC component library. Based on the modified IFC component library, a format conversion is performed to obtain an OBJ component library, and a node mapping between the OBJ component library and the information library is established. A global-local combined skeleton model is extracted from the known high-speed railway infrastructure BIM model. Multi-level semantic constraint rules are constructed based on the skeleton model. Based on the OBJ component library, the information library, the node mapping between the OBJ component library and the information library, the skeleton model, and the multi-level semantic constraint rules, the BIM model is mapped to the GIS scene and updated in real time. This invention is used for accurate virtual-real mapping of high-speed railway infrastructure BIM-GIS.

Description

Technical Field

[0001] A multi-level semantically constrained BIM-GIS virtual-real mapping method and system for high-speed railway infrastructure is used for BIM-GIS virtual-real mapping of high-speed railway infrastructure, belonging to the field of railway intelligent technology. Background Technology

[0002] Railways are the backbone of the comprehensive transportation system, a vital support for building a modern economic system, and a pioneering field in the comprehensive construction of a modern socialist country. Railway technological innovation is an important component of the national science and technology innovation system and the primary driving force for railway development. Deeply integrating modern information technologies, such as artificial intelligence, big data analytics, and the Internet of Things, into high-speed railway operation and maintenance practices provides new ideas and tools for the operation and maintenance of high-speed railways, enabling more effective management, monitoring, and maintenance of railway infrastructure.

[0003] The integration of BIM and GIS technologies offers significant advantages and complementarity in the operation and maintenance management of high-speed railway infrastructure, essentially constructing an integrated system of macro and micro data. GIS technology, with its powerful spatial analysis and query capabilities, fills the gaps in BIM's coverage of the surrounding macro environment, providing comprehensive macro-level decision support for BIM management and driving the application of BIM from single-model design to multi-dimensional spatial data. The use of GIS technology enables precise location of high-speed railway infrastructure over large areas and displays the surrounding natural and geological environment. Simultaneously, high-precision BIM models become a key source of micro-level data for GIS. By accurately locating the high-speed railway infrastructure, the high-precision 3D BIM model allows for comprehensive management, storage, display, and analysis of sensor-collected data. This integration provides real-time and detailed information for infrastructure operation and maintenance, effectively addressing various challenges. Therefore, researching BIM and GIS methods for virtual-real mapping of high-speed railway infrastructure can improve the efficiency of high-speed railway infrastructure management. High-speed railway managers can simultaneously grasp the overall system status and local details, while real-time model updates greatly improve operation and maintenance efficiency, achieving real-time and comprehensive perception of the infrastructure's operational status.

[0004] However, the existing technology has the following technical problems:

[0005] 1. When integrating BIM and GIS, there is a loss or redundancy of semantic information, which leads to low accuracy, incomplete information, and increased complexity in the mapping process.

[0006] 2. When importing a BIM model into a GIS scene, there is a terrain mismatch. The model and the terrain cannot be properly aligned, resulting in visual discrepancies and affecting the acquisition of spatial information.

[0007] 3. BIM model data is large in volume, and directly importing it into a GIS scene will lead to problems such as slow speed, low efficiency, and screen lag. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-level semantic constraint BIM-GIS virtual-real mapping method and system for high-speed railway infrastructure, which solves the problems of semantic information loss or redundancy in the existing BIM-GIS integration, resulting in low accuracy, incomplete information, and increased complexity after mapping.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0010] A multi-level semantically constrained BIM-GIS virtual-physical mapping method for high-speed railway infrastructure includes the following steps:

[0011] Step 1: Based on the acquired BIM model of the high-speed railway infrastructure, establish an IFC component library and information library stored in a tree structure, and establish the mapping relationship between the IFC component library and information library;

[0012] Step 2: Based on the filtering rules, filter, remove, and update the IFC component library to obtain the modified IFC component library;

[0013] Step 3: Traverse the tree structure of the modified IFC component library, convert each node component into OBJ format to form an OBJ component library, store it in the tree structure of the modified IFC component library, and establish a node mapping between the OBJ component library and the modified IFC component library and the information library.

[0014] Step 4: Extract the line skeleton based on the known BIM model of high-speed railway infrastructure, and extract the skeleton model of the basic structure of high-speed railway infrastructure based on the line skeleton, and perform global and local fusion to obtain a global-local combined skeleton model.

[0015] Step 5: Construct multi-level semantic constraint rules based on a skeleton model that combines global and local elements;

[0016] Step 6: Based on the modified IFC component library, OBJ component library, information library, node mapping between the OBJ component library and the modified IFC component library and information library, global-local combined skeleton model, and multi-level semantic constraint rules, the BIM model is mapped to the GIS scene, and finally the GIS scene is obtained and updated in real time.

[0017] Furthermore, the specific steps of step 1 are as follows:

[0018] Step 1.1: Convert the acquired BIM model of the high-speed railway infrastructure into an IFC file to achieve a unified BIM model format;

[0019] Step 1.2: Based on the component splitting rules, call the retrieval package in the BIMServer API to split the BIM model after it has been converted into an IFC file into components. After splitting, multi-level components are obtained and stored according to the component splitting rules to obtain the IFC component library. The component splitting rules refer to splitting the components in the IFC file by profession and splitting them into a tree structure for each profession. The professions include roadbed, bridge, tunnel and track. The tree structure includes the project as the first-level root node and the roadbed, bridge, tunnel and track as the second-level child nodes.

[0020] Step 1.3: Based on the IFC component library, perform digital-model separation to construct an information database and establish a mapping relationship between the IFC component library and the information database. The specific steps are as follows:

[0021] Step 1.31: Use the depth-first search algorithm to traverse the IFC component library, and then use the IfcOpenShell open-source library to parse the traversed IFC files, extract the global ID, geometric information and non-geometric information of the BIM model components, and save the geometric information of the components in a CSV file and the non-geometric information of the components in a JSON file. The global ID of the components needs to be stored in the CSV file and the JSON file respectively. The geometric information includes geometric modeling information and coordinate transformation information, and the non-geometric information includes semantic information and attribute information.

[0022] Step 1.32: Based on the tree structure of the IFC component library, establish a tree storage structure that corresponds one-to-one with the tree structure of the IFC component library for geometric information and non-geometric information. That is, organize the CSV file storing geometric information and the JSON file storing non-geometric information into tree storage structures that correspond one-to-one with the tree structure of the IFC component library, resulting in multiple information databases including the tree structure. The multiple information databases include the geometric information database and the non-geometric information database. At the same time, the global ID of the component in the geometric information database and the non-geometric information database is used as the node identifier of the tree structure.

[0023] Step 1.33: Based on the set mapping constraint rules, that is, to construct the mapping relationship rules between the IFC component library and each information library, the mapping constraint rules mean that the mapping between the IFC component library and the corresponding information library is injective and the tree structure is preserved. That is, each element in the tree structure of the IFC component library can only be mapped to a unique element in the tree structure of the corresponding information library, and the tree structure must be preserved. Preserving the tree structure means that if two nodes are in a parent-child relationship in the tree structure of the IFC component library, the mapping in the tree structure of the information library should also maintain the parent-child relationship.

[0024] Step 1.34: Using the mapping constraint rules as mapping constraints and the global ID as the key, establish a one-to-one mapping relationship between the IFC component library and each information library, that is, establish a one-to-one mapping relationship between the IFC component library, the geometric information library, and the non-geometric information library.

[0025] Furthermore, the specific steps of step 2 are as follows:

[0026] Step 2.1: The given filtering rules are to filter the components in the IFC component library by specialty and by node. Specialty refers to the division of high-speed railway infrastructure into different professional fields by the second-level nodes in the IFC component library. The second-level nodes in the IFC component library are selected one by one as the root node to start the specialty filtering. Based on the specialty filtering, nodes that have no functional differences but only geometric size differences are filtered. That is, the leaf nodes under the roadbed, bridge, tunnel and track specialties are filtered. For the roadbed, all leaf nodes contained in the fourth-level nodes are filtered. For bridge, tunnel and track, all leaf nodes contained in the fifth-level nodes are filtered.

[0027] Step 2.2: Modify the IFC component library based on the geometric information database and filtering rules. This involves filtering and comparing the CSV files in the geometric information database by specialty and node: Using the second-level nodes of the geometric information database as the root node, traverse each specialty. Use a depth-first search algorithm to traverse all leaf nodes under each second-level node. Compare the geometric information of each leaf node with the remaining leaf nodes. If there are differences, retain the data of that leaf node in the remaining leaf nodes. If they are the same, use the mapping relationship between the geometric information database and the IFC component library to retrieve the corresponding leaf node in the IFC component library that has the same geometric information as the one in the geometric information database. Replace the data of this leaf node with the path information of the leaf node being compared in the IFC component library. The replaced leaf node will no longer participate in subsequent comparisons, resulting in the modified IFC component library.

[0028] Furthermore, the specific steps of step 3 are as follows:

[0029] Step 3.1: Iterate through the modified IFC component library in sequence. Use the ifcConvert tool in the IfcOpenShell open source library to convert the IFC file into OBJ format file for the leaf nodes that have not been replaced. For the leaf nodes that have been replaced in the modified component library, only the path information is available. The corresponding leaf nodes only need to retain their path information. At the same time, store the global ID of the components lost during the conversion process in the OBJ file.

[0030] Step 3.2: Organize the obtained OBJ format files and leaf nodes with only path information into a tree structure identical to the modified IFC component library, forming a one-to-one OBJ component library and storing it in the database. At the same time, use the global ID of the component in the OBJ component library as the node identifier of the tree structure.

[0031] Step 3.3: Based on the mapping constraint rules and global ID, establish a one-to-one mapping relationship between the OBJ component library and the modified IFC component library, geometric information library and non-geometric information library.

[0032] Furthermore, the specific steps of step 4 are as follows:

[0033] Step 4.1: Use BIM software to manually or automatically extract the centerline and various parameters of the horizontal and vertical profiles of the known high-speed railway infrastructure BIM model. The centerline is the overall skeleton model or the line skeleton, which is an actual line with a geometric shape. The various parameters of the horizontal and vertical profiles are the attribute information of the centerline, including the starting slope information, elevation information and the location of the structures.

[0034] Step 4.2: Extract the skeleton models of four types of basic structures based on the known BIM model of high-speed railway infrastructure. These four types of basic structures are roadbed, bridges, tunnels, and tracks. The specific steps are as follows:

[0035] Step 4.21: Artificially analyze the core characteristics of the four types of basic structures: roadbed, bridge, tunnel and track, and construct abstract expression rules for the connection relationship between the components of the four types of basic structures. The abstract expression forms of the abstract expression rules include points, lines, surfaces, volumes and parameters.

[0036] Step 4.22: Based on the core features of the four types of basic structures, the components of the four types of basic structures on the centerline of the line are abstracted and expressed according to the abstract expression rules to obtain the skeleton model of the four types of basic structures, that is, the skeleton model of the basic structure of the high-speed railway infrastructure. At the same time, a global ID is assigned to the components in the skeleton model.

[0037] Step 4.3: Using the centerline of the line as the core, integrate the skeleton models of the roadbed, bridges, tunnels and tracks with the centerline of the line according to their location information to form a global-local skeleton model.

[0038] Furthermore, the specific steps of step 5 are as follows:

[0039] Step 5.1: Construct spatial layout semantic constraint rules. First, define spatial layout constraint rules to constrain the horizontal layout of the BIM model of the component along the centerline of the route. Then, use the XML Schema specification language to standardize the spatial layout constraint rules to form spatial layout semantic constraint rules. Finally, use the spatial layout semantic constraint rules to constrain the horizontal layout of the BIM model of the component.

[0040] Step 5.2: Construct spatial topological semantic constraint rules. First, define spatial topological constraint rules to constrain the combination, connection, and vertical distribution relationships of components within the BIM model of the component. Then, use the XMLSchema specification language to standardize and express the spatial topological constraint rules to form spatial topological semantic constraint rules. Finally, use the spatial topological semantic constraint rules to constrain the combination, connection, and vertical distribution relationships within the BIM model of the component, so that when the BIM model of each component is mapped, the internal topological relationships of each component's BIM model conform to the correct topological relationships, achieving continuity in the vertical direction.

[0041] Step 5.3: Construct spatial attitude semantic constraint rules. First, define spatial attitude constraint rules to constrain the attitude of the BIM model of the component itself. Then, use XML Schema to standardize the spatial attitude constraint rules to form spatial attitude semantic constraint rules. Finally, use the spatial attitude semantic constraint rules to constrain the position and attitude of the component model itself, so that when the BIM model of each component is mapped, the BIM model of each component has the correct attitude, thus achieving accurate mapping.

[0042] Furthermore, the specific steps of step 6 are as follows:

[0043] Step 6.1: Construct preliminary terrain, that is, load the DEM data obtained by the GIS platform, use the GIS platform to convert the DEM data into contour lines, then convert the contour lines into TIN models, and use the obtained TIN models as the basic terrain models.

[0044] Step 6.2, BIM model mapping, involves first placing the global-local skeleton model into the basic terrain model, then instantiating the model based on the global-local skeleton model and the global IDs of the components. The corresponding components in the OBJ component library and information library are located according to the global IDs of the components in the global-local skeleton model, and the components corresponding to the global IDs are instantiated. Finally, the instantiated model is checked and fine-tuned using spatial layout semantic constraint rules, spatial topology semantic constraint rules, and spatial pose semantic constraint rules to obtain the BIM model mapped to the GIS scene.

[0045] Step 6.3, Terrain Adaptation

[0046] In the tunnel section, the BIM model mapped to the GIS scene is compared with the basic terrain model by performing a Boolean difference operation, the basic terrain model of the tunnel section is removed, and an adaptive terrain model of the tunnel section is generated.

[0047] In roadbed or bridge sections, there is a floating situation between the BIM model mapped to the GIS scene and the basic terrain model. The basic terrain model is locally stretched according to the bottom elevation of the roadbed, pier, or abutment to generate an adaptive terrain model for the floating roadbed or bridge section. Then, the BIM model mapped to the GIS scene is inlaid with the adaptive terrain model. That is, the boundary of the BIM model mapped to the GIS scene is used to achieve the fit with the adaptive terrain model, generating the final adaptive terrain model for the floating roadbed or bridge section.

[0048] In roadbed or bridge sections, there may be instances where the basic terrain model obscures the BIM model mapped to the GIS scene. By performing a Boolean difference operation between the BIM model mapped to the GIS scene and the basic terrain model, the portion of the basic terrain model that obscures the BIM model mapped to the GIS scene is removed, generating an adaptive terrain model for roadbed / bridge sections under obscuration conditions.

[0049] Through the above steps, the basic terrain model achieves a close fit with the BIM model mapped into the GIS scene;

[0050] Step 6.4: Terrain texturing, which involves overlaying the digital orthophoto as a realistic terrain texture onto the final adaptive terrain model obtained in step 6.3 using a mapping method to obtain a realistic BIM+GIS scene.

[0051] Step 6.5, Real-time Update: When an updated component is obtained, it is indexed, located, and updated in the modified IFC component library, information library, and OBJ component library based on the component's global ID. Then, the skeleton model is locally updated based on the component's global ID. That is, for the parts of the skeleton model that need to be modified in the combination of global and local updates, Step 6.2 is executed again to update the GIS scene.

[0052] A multi-level semantically constrained BIM-GIS virtual-physical mapping system for high-speed railway infrastructure includes:

[0053] Mapping Module: Based on the acquired BIM model of high-speed railway infrastructure, establish an IFC component library and information library stored in a tree structure, and establish the mapping relationship between the IFC component library and information library;

[0054] Component modification module: Based on the filtering rules, the IFC component library is filtered, removed and updated to obtain the modified IFC component library;

[0055] Node mapping module: Traverses the tree structure of the modified IFC component library, converts each node component into OBJ format to form an OBJ component library, stores it in the tree structure of the modified IFC component library, and establishes a node mapping between the OBJ component library and the modified IFC component library and the information database.

[0056] Skeleton Model Extraction Module: Extracts the line skeleton based on the known BIM model of high-speed railway infrastructure, and extracts the skeleton model of the basic structure of high-speed railway infrastructure based on the line skeleton, and performs global and local fusion to obtain a global-local combined skeleton model;

[0057] Multi-level semantic constraint rule construction module: Constructs multi-level semantic constraint rules based on a skeleton model that combines global and local approaches;

[0058] GIS Scene Mapping and Update Module: Based on the modified IFC component library, OBJ component library, information library, node mapping between the OBJ component library and the modified IFC component library and information library, global-local combined skeleton model, and multi-level semantic constraint rules, the BIM model is mapped to the GIS scene, and finally the GIS scene is obtained and updated in real time.

[0059] Compared with the prior art, the advantages of the present invention are as follows:

[0060] I. This invention separates the digital model from the physical model, constructs a component library, a geometric information library, and a non-geometric information library, and establishes a mapping relationship. This solves the problem of semantic information loss or redundancy in the existing technology when integrating BIM and GIS, thereby improving the accuracy of the mapping and ensuring the integrity of the information. At the same time, it reduces the complexity of the mapping process and avoids the problems of slow speed, low efficiency, and screen lag caused by the large volume of BIM model data when directly imported into the GIS scene.

[0061] Second, this invention reconstructs GIS terrain and uses Boolean operations, stretching, mosaicking and other methods to solve the problem of mismatch between BIM model and GIS terrain in the existing technology, so as to achieve correct fit between BIM model and GIS terrain and avoid visual deviation.

[0062] Third, this invention utilizes an overall skeleton model for positioning, with the basic structure skeleton model serving as the BIM component mapping framework and multi-level semantic constraint rules as mapping conditions, achieving accurate BIM-GIS mapping, ensuring correct topological relationships and precise component connections in the BIM model, and enabling real-time model updates based on the global ID of the components; furthermore, by using a simplified and filtered OBJ component library as the model foundation, it enables component reuse in the BIM model, saving storage resources. Attached Figure Description

[0063] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is a general technical roadmap for the present invention;

[0065] Figure 2 This is a schematic diagram of the component splitting structure and tree-like organizational structure in this invention;

[0066] Figure 3 This is a schematic diagram of the roadbed framework in this invention;

[0067] Figure 4 This is a schematic diagram of the bridge frame in this invention;

[0068] Figure 5 This is a schematic diagram of the tunnel frame in this invention;

[0069] Figure 6 This is a schematic diagram of the track skeleton in this invention. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] A multi-level semantically constrained BIM-GIS virtual-physical mapping method for high-speed railway infrastructure includes the following steps:

[0072] Based on the acquired BIM model of high-speed railway infrastructure, an IFC component library and information library are established in a tree structure, and a mapping relationship is established between the IFC component library and the information library; the specific steps are as follows:

[0073] (1) Separation of digital and analog models

[0074] One of the core features of a BIM model is its geometric and non-geometric information. Therefore, managing the geometric and non-geometric information of a BIM model is crucial for its subsequent application.

[0075] The geometric and non-geometric information of components are associated and separated to form a unified component library as the basis for subsequent updates of component and attribute information; at the same time, the separation of digital and model information is conducive to the reuse of components.

[0076] IFC file export

[0077] Because BIM model formats differ across platforms, the IFC standard allows exporting BIM models from different platforms as IFC files, thus unifying the BIM model format. This is used to convert the acquired BIM model of high-speed railway infrastructure into an IFC file, achieving a unified BIM model format.

[0078] ② Component decomposition

[0079] a) Set component splitting rules

[0080] Component splitting rules refer to the professional splitting of components in an IFC file, and then creating a tree-like structure for each professional. Professionals include roadbed, bridges, tunnels, and tracks, such as... Figure 2 As shown, the tree structure includes projects as first-level root nodes, and roadbeds, bridges, tunnels, and tracks as second-level child nodes.

[0081] b) Component Disassembly

[0082] The BIM Server API's retrieval package is called to retrieve the entire BIM model. All objects in the IFC file obtained in step ① are traversed and split according to the model splitting rules set in the component splitting rules. The splitting results in a single smaller IFC file for export. In other words, the BIM Server API's retrieval package is called based on the component splitting rules to split the BIM model after it has been converted into an IFC file. After splitting, multi-level components are obtained, which results in multiple single smaller IFC files.

[0083] c) Establish IFC component library

[0084] The obtained IFC files are organized and stored using a tree structure according to the storage rules described in the component splitting rules, forming an orderly and well-structured IFC component library, such as... Figure 1 As shown, the IFC component library is obtained by storing the multi-level components obtained after splitting according to the component splitting rules.

[0085] For the roadbed, at the third-level node, components are classified according to their type, and different types of components such as the subgrade surface layer and the subgrade bottom layer are separated. At the fourth-level node, the model components are separated sequentially according to the component type of the third-level node and the mileage order.

[0086] For bridges, at the third-level node, each bridge is divided into sections from the starting point to the end point according to the route mileage; at the fourth-level node, each bridge is divided into superstructure, substructure, supports, and auxiliary structures according to its structural characteristics; at the fifth-level node, the model components are sequentially divided into sections from the starting point to the end point according to the structural classification of the fourth-level node.

[0087] For tunnels, at the third-level node, each tunnel is divided into sections from the starting point to the end point according to the route mileage; at the fourth-level node, each tunnel is divided into portal structure, tunnel body structure, auxiliary structure, and drainage system according to its structural characteristics; at the fifth-level node, the model components are sequentially divided into sections from the starting point to the end point according to the structural classification of the fourth-level node.

[0088] For the track, at the third-level node, it is classified into bridge sections, roadbed sections, and tunnel sections according to the type of structures under the track. Each section is then divided sequentially from the starting point to the end point according to the line mileage. At the fourth-level node, each section is divided into rails, rail joints, fasteners, sleepers, turnouts, rail expansion joints, ballast, reinforcing equipment, and auxiliary equipment. At the fifth-level node, the components are then divided sequentially from the starting point to the end point.

[0089] ③ Separation of digital and analog models

[0090] This information management method, based on the separation of digital and physical components in BIM models, enables unified management of components. Its core idea is to associate and separate each component in the building model with its related information, forming a unified component library. Through this separation of digital and physical component management, refined management of components can be achieved, ensuring information consistency and accuracy, thereby improving the efficiency and quality of building projects.

[0091] Furthermore, since non-geometric information is lost when converting from IFC format to OBJ format, it is essential to perform digital-model separation to preserve the non-geometric and semantic information of the BIM model.

[0092] The following steps are taken to construct an information database based on the IFC component library by separating the digital model from the data model, and to establish a mapping relationship between the IFC component library and the information database:

[0093] a) Parse the IFC file and extract geometric and non-geometric information.

[0094] The geometric and non-geometric information of the BIM model itself is of great significance for coordinate transformation, visual analysis, and deformation analysis. A depth-first search algorithm is used to traverse the IFC component library. Then, the IfcOpenShell open-source library is used to parse the traversed IFC files, extracting the global ID, geometric information, and non-geometric information of the BIM model. Geometric information, including geometric modeling and coordinate transformation information, is stored in a CSV file; non-geometric information, including semantic and attribute information, is stored in a JSON file. The global ID of the component is stored in both the CSV and JSON files.

[0095] b) Establish a storage structure for information files to form an information database.

[0096] Based on the tree-like storage structure of the IFC component library, the obtained geometric and non-geometric information are used to establish a tree-like storage structure that corresponds one-to-one with the IFC component library. Figure 2 The tree structure of the IFC component library corresponds one-to-one with the geometric information library and the non-geometric information library. Geometric information and non-geometric information are stored in the database to obtain multiple information libraries, including the tree-structured geometric information library and the non-geometric information library. At the same time, the global ID of the component in the geometric information library and the non-geometric information library is used as the node identifier of the tree structure.

[0097] c) Establish a mapping relationship between the IFC component library and the information library.

[0098] Because the ID information of components may be lost after converting IFC files to OBJ files, causing the components to be unable to be associated with information, a mapping relationship is established to achieve the association between the model and the information. The specific steps are as follows:

[0099] First, mapping constraints are defined to ensure that the mapping between the IFC component library and each information library is injective and maintains the tree structure. This means that each element in the first tree can only be mapped to a unique element in the second tree, and the tree structure must be preserved (i.e., if two nodes are parent-child in the first tree, their mapping in the second tree should also maintain this parent-child relationship). Specifically, the mapping constraints mean that the mapping between the IFC component library and the corresponding information library is injective and maintains the tree structure. This means that if two nodes are parent-child in the IFC component library's tree structure, their mapping in the information library's tree structure should also maintain this parent-child relationship.

[0100] Secondly, a depth-first search algorithm is used to traverse the tree-structured IFC component library, the tree-structured geometric information library, and the tree-structured non-geometric information library to obtain their nodes.

[0101] Finally, using mapping constraint rules as mapping constraints and global ID as the key, a one-to-one mapping relationship is established between the IFC component library and each information library, that is, a one-to-one mapping relationship is established between the IFC component library, the geometric information library, and the non-geometric information library.

[0102] (2) Implementation of component reuse

[0103] The issue of BIM component reuse arises because railway BIM models contain many standardized components, resulting in a large number of identical geometric components in the IFC component library, which consumes significant storage space. Therefore, it is necessary to filter components, removing duplicates from the IFC component library to achieve component reuse. This involves filtering, removing, and updating components in the IFC component library based on filtering rules to obtain a modified IFC component library. The specific steps are as follows:

[0104] Set filtering rules

[0105] First, a selection process is conducted by major, by... Figure 2 As can be seen, the second-level nodes in the IFC component library divide high-speed railway infrastructure into different professional fields. Therefore, when screening components, it is necessary to strictly distinguish between components of different professional fields and select the second-level nodes in the IFC component library as the root nodes to start the professional screening.

[0106] Secondly, node-by-node screening is conducted. Based on the professional screening, nodes with only geometric dimensional differences and no functional differences are further screened. For example, all leaf nodes below the rail node are rail components used in railway projects; these components differ only in geometric dimensions and have no functional differences. In such cases... Figure 2 The IFC component library shown here filters components at the first level, which is the leaf node (i.e., the end node in the tree structure, which is a node without child nodes). For roadbeds, it filters all leaf nodes in the fourth level nodes. For bridges, tunnels and tracks, it filters all leaf nodes in the fifth level nodes.

[0107] The modified IFC component library is obtained after component filtering, removal, and updating.

[0108] A geometric information database is selected as the basis for component selection or comparison. Selection is performed according to selection rules, specifically by subject and node-based comparison of the CSV files in the geometric information database. Specifically, the second-level nodes of the geometric information database are used as the root nodes, and the database is traversed one by one by subject using a depth-first search algorithm to reach all leaf nodes under each second-level node. The geometric information of each leaf node is compared with the remaining leaf nodes. If there are differences, the data in the compared leaf node is retained; if they are the same, the mapping relationship between the geometric information database and the IFC component database is used to retrieve the corresponding leaf node in the IFC component database. The data of the corresponding leaf node in the IFC component database is replaced with the path information of the compared leaf node in the IFC component database. Simultaneously, the replaced leaf node is no longer included in subsequent comparisons, resulting in the modified IFC component database.

[0109] The core of this step is to compare the geometric information in the geometric information database. For leaf nodes with the same geometric information, the mapping relationship is used to find the corresponding leaf node in the IFC component library. The data of the leaf node is modified to be the path information of the node used as a reference for comparison in the IFC component library. At the same time, the modified leaf node does not participate in the subsequent comparison.

[0110] Modifying the IFC component library using the above traversal method and updating the IFC component library yields the modified IFC component library. This step enables component reuse and builds a standardized and lightweight component library.

[0111] (3) Data format conversion IFC-OBJ

[0112] BIM-GIS virtual-physical mapping requires unifying BIM and GIS data onto a single platform where both types of data can be displayed compatiblely. This involves BIM data format conversion. BIM entity models need to be converted into GIS surface models to achieve seamless integration of BIM models and 3D GIS scenes. Simultaneously, the entire operation and maintenance process can be simulated in a virtual geographic environment, enabling unified and integrated management of railway engineering information. The main process involves first converting different BIM model formats into IFC files, and then converting the IFC files into OBJ files. The conversion of different BIM model formats to IFC files has already been implemented; therefore, the following steps address the issue of converting IFC files to OBJ files.

[0113] Based on the previous steps, the components for BIM model format conversion in this step are the components in the modified IFC component library. If there are errors / mistakes in the data after format conversion, the components can be modified in a targeted manner to improve accuracy. In addition, based on the obtained standard and lightweight IFC component library, the amount of data for format conversion in this step is greatly reduced.

[0114] This step uses the method in IfcOpenShell for model format conversion. The key is to traverse the modified tree-structured IFC component library to form a corresponding tree-structured OBJ component library. Furthermore, by utilizing the component's global ID, a one-to-one mapping can be achieved between the OBJ component library and the modified IFC component library, geometric information library, and non-geometric information library. At the same time, the attributes, semantics, and other non-geometric information lost by the components after format conversion to OBJ components can also be re-established through the established mapping relationship.

[0115] Compared to existing technologies, this step, by establishing a tree-structured OBJ component library, not only achieves a one-to-one mapping with the modified IFC component library but also associates the OBJ component library with various information repositories, realizing the association between OBJ format components and their attributes and semantic information. IFC-OBJ sequentially traverses the modified IFC component library, converting unreplaced leaf nodes to OBJ format files using the ifcConvert tool from the IfcOpenShell open-source library. For replaced leaf nodes in the modified component library, only path information is provided; the corresponding leaf nodes must retain their path information. The component's global ID, serving as the key for linking the component library and information repositories, also needs to be stored in the OBJ file (i.e., storing the global IDs of components lost during the conversion process in the OBJ file).

[0116] Establish OBJ component library

[0117] The obtained OBJ format files and leaf nodes containing only path information are organized into a tree structure identical to that of the IFC component library and stored in the database, forming an OBJ component library with a one-to-one tree structure. Furthermore, the global ID of each component serves as the node identifier in the OBJ component library within the tree structure.

[0118] Establish a mapping relationship between the OBJ component library and various information databases.

[0119] Because non-geometric information is lost after converting IFC to OBJ, it is necessary to establish a mapping relationship between the OBJ component library and the information library. Using the component's global ID as the key, a mapping relationship is established between the OBJ component library and the modified IFC component library, the geometric information library, and the non-geometric information library, following the same method used to establish the mapping relationship between the IFC component library and the information library. This enables the association between components in the OBJ component library and components in the modified IFC component library, as well as the geometric and non-geometric information of the components.

[0120] (4) Skeleton model extraction

[0121] Extracting the overall framework of a high-speed railway line from an existing BIM model of high-speed railway infrastructure involves extracting the line centerline, overall framework model, or line skeleton. The core idea is to use a series of points, lines, surfaces, volumes, and parameters to describe the BIM model and outline the key features of the high-speed railway project. The core of this step is using the skeleton model for line and component positioning, and then instantiating the component models in the GIS scene based on the positioning information. Besides using skeleton elements as input for shaping and positioning, the component models can maintain constraints based on the skeleton model structure; modifications to the skeleton will drive automatic updates to the model's geometry and position.

[0122] ① Overall skeleton model extraction, i.e., line centerline extraction

[0123] The overall skeleton model consists of the centerlines of each route. Extracting the overall skeleton model here means extracting the route centerlines. Using the functions provided by BIM software, the route centerlines and various parameters of the horizontal and vertical profiles are extracted manually or automatically, including the initial slope information, elevation information, and the location of structures.

[0124] This step ultimately yields the centerline framework and key parameter information of the high-speed railway infrastructure. This overall framework can be used as a reference system for locating the components of high-speed railway infrastructure projects, namely, the subsequent location of the four types of line foundation structures: railway subgrade, bridges, tunnels, and tracks.

[0125] ② Extraction of the skeleton model of high-speed railway infrastructure basic structures

[0126] Taking roadbed, bridge, tunnel and track as examples, the skeleton models of these four types of basic structures are extracted respectively.

[0127] a) Artificially analyze the core characteristics of four types of basic structures: roadbed, bridge, tunnel and track, and construct abstract expression rules for the connection relationship between the components of the four types of basic structures. The abstract expression forms include points, lines, surfaces, volumes and parameters.

[0128] like Figure 3 As shown, the roadbed comprises three core components: the subgrade surface layer, the subgrade bottom layer, and the embankment below the subgrade. The key characteristic of these components is their continuity in the transverse direction and their height in the longitudinal direction. Therefore, components that are continuous in the transverse direction and have height in the longitudinal direction are represented by lines, the transverse connection relationships between similar components are represented by points, and the thickness in the longitudinal direction is represented by dashed lines. Specifically, taking the subgrade surface layer as an example, the starting and ending points of the top of the subgrade surface layer are extracted along the route direction. The starting and ending points are taken as the midpoints of the component cross-sections, and their coordinate information is recorded. The same applies to the subgrade bottom layer and the embankment below the subgrade.

[0129] like Figure 4As shown, a bridge includes core components such as beams, bearings, pad stones, piers / aprons, pile caps, and pile foundations. Among them, the beams and abutment surfaces have similar characteristics to roadbeds, and their components are abstracted as having endpoints, starting points, and ending points. The centroids of the slopes, bearings, and pad stones are extracted and abstracted as points, and the connection relationships between the slopes and abutments, and between bearings and pad stones are expressed by dashed lines. The pier caps, piers or abutments, pile caps, and pile foundations have continuity in the longitudinal direction and are abstracted as lines with length (consistent with the length of the entity), and the connection relationships in the middle are expressed by points, while the connection relationship between pile foundations and pile caps is expressed by dashed lines.

[0130] like Figure 5 As shown, the core structure of the tunnel is the tunnel body, which includes components such as initial support, anchor bolts, steel frames, secondary lining, initial support invert, secondary lining invert, and invert filling. It also includes ancillary facilities such as cable trenches and central drainage ditches. The initial support, initial support invert, secondary lining, secondary lining invert, and invert filling have widths along the tunnel route; the starting cross-section of these components is extracted to represent them. The steel frame is represented by its cross-section and top midpoint. For slender cylindrical components like anchor bolts, the centers of their top and bottom surfaces are extracted as endpoints, and the connection points are abstracted as line entities. In complex tunnels, the number of anchor bolts is enormous; therefore, the endpoints where they connect to the tunnel body are extracted and their attitude parameters are recorded for subsequent mapping.

[0131] like Figure 6 As shown, the track mainly consists of core components such as ballastless track bed, rails, sleepers, and fasteners. The track bed, rails, and subgrade share similar characteristics of being continuous laterally and having thickness longitudinally. Therefore, these track components are represented as lines, the transverse connections between similar components are represented as points, and the longitudinal height is represented by dashed lines. The track bed and subgrade components are represented in the same way; for rail components, the starting and ending points of the portion where the bottom connects to the track slab are extracted along the track direction, with the starting and ending points taken as the midpoints of the rail cross-section and their positioning information recorded; sleepers are discontinuous laterally but have a regular distribution, and their endpoints are extracted and connected as line entities for abstract representation; the fasteners connecting sleepers and rails, as well as the joints between rails, are extracted and represented as point entities.

[0132] b) Extraction of skeleton models for four types of basic structures

[0133] Based on the core characteristics of the four types of basic structures, the components of these four types of basic structures along the centerline of the railway line are abstracted and expressed according to abstraction rules, resulting in the skeleton models of the four types of basic structures, i.e., the skeleton models of the basic structures of high-speed railway infrastructure. Simultaneously, global IDs are assigned to the components in the skeleton models. Through this step, the skeleton models of the roadbed, bridges, tunnels, and tracks are extracted. That is, using the simplest points, lines, surfaces, volumes, and parameters, the road, bridges, tunnels, and tracks are expressed as skeleton structures with simple shapes, precise coordinates, and clear topology.

[0134] ③ A skeleton model that combines "global and local" approaches, i.e., achieving skeleton fusion.

[0135] The overall skeleton model is used as the basis to constrain the layout of the four core elements on the overall skeleton model. With the center line of the line as the core, the skeleton models of the roadbed, bridges, tunnels and tracks are integrated with the center line of the line according to the location information to form a global-local skeleton model.

[0136] (5) Construct multi-level semantic constraint rules for subsequent accurate mapping (processing of spatial relationships, determination of mapping rules).

[0137] Precise control of component positions, combination relationships, and connection relationships is crucial for accurate model mapping. Therefore, relevant constraints are established for model checking and correction during subsequent instantiation mapping based on the global-local skeleton model. The overall skeleton model and local skeletons for roads, bridges, tunnels, and tracks are obtained. Based on the global-local skeleton model, corresponding components are mapped to abstract representations of the skeletons. Then, the instantiation process of replacing the skeleton model with the component model is completed using the component's global ID. However, mapping component models solely based on skeleton information may lead to errors such as topological misalignment. Therefore, a multi-level semantic constraint library is constructed to constrain the component model mapping. This not only ensures that the component model is instantiated according to the correct mapping rules but also allows for error checking after mapping. Furthermore, when updating component models, the overall skeleton model is updated first, followed by component model replacement. Semantic constraint rules are also needed to constrain and check this process, ensuring efficient and accurate model updates.

[0138] Constructing semantic constraint rules for spatial layout

[0139] Spatial layout involves the overall control of layout rules, including the close connection and continuity of tracks, roadbeds, bridges, and tunnels in spatial layout; the continuity of roadbeds in non-bridge / tunnel sections; and the spatial continuity of tracks without interruption. Furthermore, the spatial layout of the component units of various main bodies is also closely interconnected. Therefore, spatial layout constraints are used to standardize the overall layout when mapping high-speed railway infrastructure into BIM models.

[0140] a) Define spatial layout constraints rules

[0141] Define spatial layout constraints for the BIM model of components to constrain the lateral layout along the centerline of the line. These constraints include: the track, roadbed, bridge, and tunnel are closely connected and continuous in space; the roadbed is continuous in non-bridge / tunnel sections; the track is continuous in space and cannot be interrupted; there are road-bridge transition sections and road-tunnel transition sections in the roadbed, and precise transitions are required at both ends of the bridges and tunnels; except for the substructure and supports of bridges, laterally distributed components are not allowed to overlap or be separated from each other.

[0142] b) Standardized semantic constraints

[0143] The spatial layout constraint rules are standardized and expressed using the XML Schema specification language, forming spatial layout semantic constraint rules.

[0144] c) Using spatial layout semantic constraint rules to constrain the horizontal layout of the BIM model of components is a macroscopic mapping constraint.

[0145] ② Construct spatial topological semantic constraint rules

[0146] The components of different structures such as tracks, roadbeds, bridges, and tunnels have different topological relationships. The combination of components can form different assemblies. It is necessary to clarify the topological structure between components and use spatial topological constraints to standardize the connection and combination relationships of components when mapping the BIM model of high-speed railway infrastructure.

[0147] a) Define spatial topological constraint rules

[0148] First, define spatial topological constraint rules to constrain the combination, connection, and longitudinal distribution relationships of the internal components of the BIM model of the components. This includes ensuring that the combination relationships of internal components of roadbeds, bridges, tunnels, and tracks conform to the structural rules of the structures. For example, the roadbed consists of the subgrade surface layer, the subgrade bottom layer, and the embankment below the subgrade from top to bottom. The bridge piers consist of pile foundations, abutments, piers / abutments, bearings, pad stones, and bridge deck from bottom to top. Track fasteners are components that fix the rails to the sleepers. Therefore, in the vertical direction, the rails are located at the top, and the fasteners fix the rails to the sleepers by clamping the bottom of the rails. The sleepers are located directly below the fasteners.

[0149] b) Standardized semantic constraints

[0150] The spatial topological constraint rules are standardized and expressed using the XML Schema specification language, forming spatial topological semantic constraint rules.

[0151] c) By using spatial topological semantic constraint rules, the combination relationship, connection relationship and vertical distribution relationship of the component model are constrained, so that when the component model is mapped, the internal topological relationship of the component model conforms to the correct topological relationship and achieves continuity in the vertical direction. This belongs to the meso-level mapping constraint.

[0152] ③ Construct spatial attitude semantic constraint rules

[0153] For the task of mapping BIM models of high-speed railway infrastructure, the mapping of component models into GIS scenes needs to consider the adaptation of spatial coordinates and angles. Therefore, spatial attitude constraints are used to constrain the accuracy of the location of the BIM model mapping of high-speed railway infrastructure.

[0154] a) Define spatial attitude semantic rules

[0155] In step (1), we extracted the geometric information of the component, which included the orientation information of the component. We used this information as the orientation constraint during mapping, that is, we first defined the spatial orientation constraint rules for the orientation constraint of the BIM model of the component itself.

[0156] b) Standardized semantic constraints

[0157] The spatial pose semantic rules are standardized and expressed using the XML Schema specification language, forming spatial pose semantic constraint rules.

[0158] c) By using spatial pose semantic rules, the position and pose of the component model itself are constrained, so that when the model of each component is mapped, the pose of the model of each component is correct, thus achieving accurate mapping. This belongs to microscopic mapping constraints.

[0159] (6) Mapping BIM models to GIS scenes

[0160] ① Constructing the initial terrain

[0161] Load DEM data into the SuperMap / ArcGIS platform, convert the DEM to contour lines using the platform's function, and then convert the contour lines to a TIN model. Use the resulting TIN model as the base terrain model. This step of converting the DEM to contour lines before converting it to a TIN model captures terrain details and features more accurately than directly converting the DEM to a TIN model, especially in areas with complex terrain.

[0162] BIM model mapping

[0163] First, the global-local skeleton model is instantiated by placing it into the basic terrain model. Then, the OBJ component library is located based on the global ID of the component in the skeleton model, and the component is called. At the same time, the spatial layout semantic constraint rules, spatial topology semantic constraint rules, and spatial pose semantic constraint rules are used as input constraint components for placement and combination to obtain the BIM model mapped to the GIS scene, thus achieving accurate instantiation.

[0164] Then, based on the mapping relationship between the OBJ component library and each information library, the corresponding component information library is located, and the component information is associated with the component model to ensure the integrity of the component's geometry and attribute information.

[0165] ③ Terrain Reconstruction

[0166] In the tunnel section, the BIM model mapped to the GIS scene is compared with the basic terrain model by performing a Boolean difference operation, the basic terrain model of the tunnel section is removed, and an adaptive terrain model of the tunnel section is generated.

[0167] In roadbed or bridge sections, there is a floating situation between the BIM model mapped to the GIS scene and the basic terrain model. The basic terrain model is locally stretched according to the bottom elevation of the roadbed, pier, or abutment to generate an adaptive terrain model for the floating roadbed or bridge section. Then, the BIM model mapped to the GIS scene is inlaid with the adaptive terrain model. That is, the boundary of the BIM model mapped to the GIS scene is used to achieve the fit with the adaptive terrain model, generating the final adaptive terrain model for the floating roadbed or bridge section.

[0168] In roadbed or bridge sections, there may be instances where the basic terrain model obscures the BIM model mapped to the GIS scene. By performing a Boolean difference operation between the BIM model mapped to the GIS scene and the basic terrain model, the portion of the basic terrain model that obscures the BIM model mapped to the GIS scene is removed, generating an adaptive terrain model for roadbed / bridge sections under obscuration conditions.

[0169] Through the above steps, the basic terrain model achieves a perfect fit with the BIM model mapped into the GIS scene.

[0170] ④ Terrain texture

[0171] The digital orthophoto is used as a terrain texture (it is necessary to ensure that the spatial coordinates of the digital orthophoto and the TIN model are consistent; if they are inconsistent, a projection conversion needs to be performed on the GIS platform to achieve coordinate consistency) and is overlaid on the reconstructed terrain TIN model using a texture mapping method.

[0172] ⑤ Real-time updates: When an updated component is obtained, it is indexed and located based on its global ID, and then the modified IFC component library, information library, and OBJ component library are updated. Next, the skeleton model is locally updated based on the component's global ID. Finally, BIM model mapping, terrain reconstruction, and terrain texturing are performed again on the updated portion of the combined global and local skeleton model to update the GIS scene. Specifically:

[0173] Generate a copy of the skeleton model that combines global and local data before the update;

[0174] Modifications are made on a copy of the skeleton model. For components requiring updates, the global ID of the corresponding component on the copy of the skeleton model is located, quickly indexed, and the update position is determined. The parts requiring updates are then modified on the skeleton model. The modified skeleton model is then instantiated and updated, including replacing component models and adding, deleting, or modifying attribute information. Specifically, during updates, components are entered in different BIM model formats. First, they are standardized into IFC files. Then, the component is stored in the corresponding position in the modified IFC component library, replacing the original information. Finally, the component format is converted to OBJ components and stored in the OBJ component library. When updating components, the updated components in the OBJ component library are called.

[0175] Multi-source information association: Operation and maintenance information and attribute information are associated with the BIM model through the database, and updates to the database can update the model information. The core of local component information association lies in associating multi-source information with the global ID of the corresponding component in the road, bridge, tunnel, and track skeleton model, enabling the supplementation and updating of component-level information. The core of global / section information association lies in associating multi-source information with the overall route model, enabling the supplementation and updating of section / global information.

Claims

1. A multi-level semantically constrained BIM-GIS virtual-real mapping method for high-speed railway infrastructure, characterized in that, Includes the following steps: Step 1: Based on the acquired BIM model of the high-speed railway infrastructure, establish an IFC component library and information library stored in a tree structure, and establish the mapping relationship between the IFC component library and information library; Step 2: Based on the filtering rules, filter, remove, and update the IFC component library to obtain the modified IFC component library; Step 3: Traverse the tree structure of the modified IFC component library, convert each node component into OBJ format to form an OBJ component library, store it in the tree structure of the modified IFC component library, and establish a node mapping between the OBJ component library and the modified IFC component library and the information library. Step 4: Extract the line skeleton based on the known BIM model of high-speed railway infrastructure, and extract the skeleton model of the basic structure of high-speed railway infrastructure based on the line skeleton, and perform global and local fusion to obtain a global-local combined skeleton model. Step 5: Construct multi-level semantic constraint rules based on a global-local skeleton model; the specific steps are as follows: Step 5.1: Construct spatial layout semantic constraint rules. Specifically, first define spatial layout constraint rules to constrain the horizontal layout of the BIM model of the component along the centerline of the route. Then, use the XML Schema specification language to standardize the spatial layout constraint rules to form spatial layout semantic constraint rules. Finally, use the spatial layout semantic constraint rules to constrain the horizontal layout of the BIM model of the component. Step 5.2: Construct spatial topological semantic constraint rules. Specifically, first, define spatial topological constraint rules to constrain the combination, connection, and vertical distribution relationships of components within the BIM model of the component. Then, use the XMLSchema specification language to standardize and express the spatial topological constraint rules to form spatial topological semantic constraint rules. Finally, use the spatial topological semantic constraint rules to constrain the combination, connection, and vertical distribution relationships within the BIM model of the component, so that when the BIM model of each component is mapped, the internal topological relationships of each component's BIM model conform to the correct topological relationships, achieving continuity in the vertical direction. Step 5.3: Construct spatial attitude semantic constraint rules. Specifically, first define spatial attitude constraint rules to constrain the attitude of the BIM model of the component itself, then use XML Schema to standardize the spatial attitude constraint rules to form spatial attitude semantic constraint rules, and finally use the spatial attitude semantic constraint rules to constrain the position and attitude of the component model itself, so that when the BIM model of each component is mapped, the BIM model of each component has the correct attitude, thus achieving accurate mapping. Step 6: Based on the modified IFC component library, OBJ component library, information library, node mapping between the OBJ component library and the modified IFC component library and information library, global-local combined skeleton model, and multi-level semantic constraint rules, the BIM model is mapped to the GIS scene, and finally the GIS scene is obtained and updated in real time.

2. The multi-level semantic constraint BIM-GIS virtual-real mapping method for high-speed railway infrastructure according to claim 1, characterized in that: The specific steps of step 1 are as follows: Step 1.1: Convert the acquired BIM model of the high-speed railway infrastructure into an IFC file to achieve a unified BIM model format; Step 1.2: Based on the component splitting rules, call the retrieval package in the BIMServer API to split the BIM model after it has been converted into an IFC file into components. After splitting, multi-level components are obtained and stored according to the component splitting rules to obtain the IFC component library. The component splitting rules refer to splitting the components in the IFC file by profession and splitting them into a tree structure for each profession. The professions include roadbed, bridge, tunnel and track. The tree structure includes the project as the first-level root node and the roadbed, bridge, tunnel and track as the second-level child nodes. Step 1.3: Based on the IFC component library, perform digital-model separation to construct an information database and establish a mapping relationship between the IFC component library and the information database. The specific steps are as follows: Step 1.31: Use the depth-first search algorithm to traverse the IFC component library, and then use the IfcOpenShell open-source library to parse the traversed IFC files, extract the global ID, geometric information and non-geometric information of the BIM model components, and save the geometric information of the components in a CSV file and the non-geometric information of the components in a JSON file. The global ID of the components needs to be stored in the CSV file and the JSON file respectively. The geometric information includes geometric modeling information and coordinate transformation information, and the non-geometric information includes semantic information and attribute information. Step 1.32: Based on the tree structure of the IFC component library, establish a tree storage structure with geometric information and non-geometric information corresponding one-to-one with the tree structure of the IFC component library. Specifically, organize the CSV file storing geometric information and the JSON file storing non-geometric information into tree storage structures that correspond one-to-one with the tree structure of the IFC component library, resulting in multiple information databases including the tree structure. The multiple information databases include the geometric information database and the non-geometric information database. At the same time, the global ID of the component in the geometric information database and the non-geometric information database is used as the node identifier of the tree structure. Step 1.33: Based on the set mapping constraint rules, specifically, construct the mapping relationship rules between the IFC component library and each information library. The mapping constraint rules mean that the mapping between the IFC component library and the corresponding information library is injective and the tree structure is preserved. Specifically, each element in the tree structure of the IFC component library can only be mapped to a unique element in the tree structure of the corresponding information library, and the tree structure must be preserved. Preserving the tree structure means that if two nodes are in a parent-child relationship in the tree structure of the IFC component library, the mapping in the tree structure of the information library should also maintain the parent-child relationship. Step 1.34: Using mapping constraint rules as mapping constraints and global ID as the key, establish a one-to-one mapping relationship between the IFC component library and each information library. Specifically, establish a one-to-one mapping relationship between the IFC component library, the geometric information library, and the non-geometric information library.

3. The multi-level semantic constraint BIM-GIS virtual-real mapping method for high-speed railway infrastructure according to claim 2, characterized in that: The specific steps of step 2 are as follows: Step 2.1: The given filtering rules are to filter the components in the IFC component library by specialty and by node. Specialty refers to the division of high-speed railway infrastructure into different professional fields by the second-level nodes in the IFC component library. The second-level nodes in the IFC component library are selected one by one as the root node to start the specialty filtering. Based on the specialty filtering, nodes that have no functional differences but only geometric size differences are filtered. Specifically, the leaf nodes under the roadbed, bridge, tunnel and track specialties are filtered. For the roadbed, all leaf nodes contained in the fourth-level nodes are filtered. For bridge, tunnel and track, all leaf nodes contained in the fifth-level nodes are filtered. Step 2.2: Modify the IFC component library based on the geometric information database and filtering rules. Specifically, perform professional and node-based filtering and comparison on the CSV files in the geometric information database: Take the second-level nodes of the geometric information database as the root node and traverse them one by one by professional. Use the depth-first algorithm to traverse to all leaf nodes under each second-level node. Compare the geometric information of each leaf node with the remaining leaf nodes one by one. If there are differences, retain the data of that leaf node in the remaining leaf nodes. If they are the same, the mapping relationship between the geometric information database and the IFC component database is used to retrieve the leaf node in the IFC component database that has the same geometric information as the corresponding leaf node in the geometric information database. The data of this leaf node is replaced with the path information of the leaf node being compared in the IFC component database. At the same time, the replaced leaf node will no longer participate in subsequent comparisons, and the modified IFC component database is finally obtained.

4. The multi-level semantic constraint BIM-GIS virtual-real mapping method for high-speed railway infrastructure according to claim 3, characterized in that: The specific steps of step 3 are as follows: Step 3.1: Iterate through the modified IFC component library in sequence. Use the ifcConvert tool in the IfcOpenShell open source library to convert the IFC file into OBJ format file for the leaf nodes that have not been replaced. For the leaf nodes that have been replaced in the modified component library, only the path information is available. The corresponding leaf nodes only need to retain their path information. At the same time, store the global ID of the components lost during the conversion process in the OBJ file. Step 3.2: Organize the obtained OBJ format files and leaf nodes with only path information into a tree structure identical to the modified IFC component library, forming a one-to-one OBJ component library and storing it in the database. At the same time, use the global ID of the component in the OBJ component library as the node identifier of the tree structure. Step 3.3: Based on the mapping constraint rules and global ID, establish a one-to-one mapping relationship between the OBJ component library and the modified IFC component library, geometric information library and non-geometric information library.

5. The multi-level semantic constraint BIM-GIS virtual-real mapping method for high-speed railway infrastructure according to claim 4, characterized in that: The specific steps of step 4 are as follows: Step 4.1: Use BIM software to manually or automatically extract the centerline and various parameters of the horizontal and vertical profiles of the known high-speed railway infrastructure BIM model. The centerline is the overall skeleton model or the line skeleton, which is an actual line with a geometric shape. The various parameters of the horizontal and vertical profiles are the attribute information of the centerline, including the starting slope information, elevation information and the location of the structures. Step 4.2: Extract the skeleton models of four types of basic structures based on the known BIM model of high-speed railway infrastructure. These four types of basic structures are roadbed, bridges, tunnels, and tracks. The specific steps are as follows: Step 4.21: Artificially analyze the core characteristics of the four types of basic structures: roadbed, bridge, tunnel and track, and construct abstract expression rules for the connection relationship between the components of the four types of basic structures. The abstract expression forms of the abstract expression rules include points, lines, surfaces, volumes and parameters. Step 4.22: Based on the core features of the four types of basic structures, the components of the four types of basic structures on the centerline of the line are abstracted and expressed according to the abstract expression rules to obtain the skeleton model of the four types of basic structures. Specifically, the skeleton model of the basic structure of high-speed railway infrastructure is obtained. At the same time, a global ID is assigned to the components in the skeleton model. Step 4.3: Using the centerline of the line as the core, integrate the skeleton models of the roadbed, bridges, tunnels and tracks with the centerline of the line according to their location information to form a global-local skeleton model.

6. The high-speed railway infrastructure BIM-GIS virtual-real mapping method with multi-level semantic constraints according to claim 5, characterized in that: The specific steps of step 6 are as follows: Step 6.1: Construct preliminary terrain. Specifically, load the DEM data obtained from the GIS platform, convert the DEM data into contour lines using the GIS platform, then convert the contour lines into a TIN model, and use the obtained TIN model as the basic terrain model. Step 6.2, BIM model mapping, specifically involves first placing the global-local skeleton model into the basic terrain model, then instantiating the model based on the global-local skeleton model and the global IDs of the components, locating the corresponding components in the OBJ component library and information library according to the global IDs of the components in the global-local skeleton model, and calling the component corresponding to the global ID for instantiation, and finally checking and fine-tuning the instantiated model using spatial layout semantic constraint rules, spatial topology semantic constraint rules, and spatial pose semantic constraint rules to obtain the BIM model mapped to the GIS scene; Step 6.3, Terrain Adaptation In the tunnel section, the BIM model mapped to the GIS scene is compared with the basic terrain model by performing a Boolean difference operation, the basic terrain model of the tunnel section is removed, and an adaptive terrain model of the tunnel section is generated. In roadbed or bridge sections, there is a floating situation between the BIM model mapped to the GIS scene and the basic terrain model. The basic terrain model is locally stretched according to the bottom elevation of the roadbed, pier, or abutment to generate an adaptive terrain model for the floating roadbed or bridge section. Then, the BIM model mapped to the GIS scene and the adaptive terrain model are inlaid. Specifically, the boundary of the BIM model mapped to the GIS scene is used to achieve the fit with the adaptive terrain model to generate the final adaptive terrain model for the floating roadbed or bridge section. In roadbed or bridge sections, there may be instances where the basic terrain model obscures the BIM model mapped to the GIS scene. By performing a Boolean difference operation between the BIM model mapped to the GIS scene and the basic terrain model, the portion of the basic terrain model that obscures the BIM model mapped to the GIS scene is removed, generating an adaptive terrain model for roadbed / bridge sections under obscuration conditions. Through the above steps, the basic terrain model achieves a close fit with the BIM model mapped into the GIS scene; Step 6.4: Terrain texturing. Specifically, digital orthophotos are used as terrain textures and overlaid onto the final adaptive terrain model obtained in Step 6.3 using a mapping method to obtain a realistic BIM+GIS scene. Step 6.5, Real-time Update: Specifically, when an updated component is obtained, it is indexed, located, and updated in the modified IFC component library, information library, and OBJ component library based on the component's global ID. Then, the skeleton model is locally updated based on the component's global ID. Specifically, Step 6.2 is executed again for the parts of the skeleton model that need to be modified in the combination of global and local updates to achieve the update of the GIS scene.

7. A high-speed railway infrastructure BIM-GIS virtual-real mapping system with multi-level semantic constraints, characterized in that, include: Mapping Module: Based on the acquired BIM model of high-speed railway infrastructure, establish an IFC component library and information library stored in a tree structure, and establish the mapping relationship between the IFC component library and information library; Component modification module: Based on the filtering rules, the IFC component library is filtered, removed and updated to obtain the modified IFC component library; Node mapping module: Traverses the tree structure of the modified IFC component library, converts each node component into OBJ format to form an OBJ component library, stores it in the tree structure of the modified IFC component library, and establishes a node mapping between the OBJ component library and the modified IFC component library and the information database. Skeleton Model Extraction Module: Extracts the line skeleton based on the known BIM model of high-speed railway infrastructure, and extracts the skeleton model of the basic structure of high-speed railway infrastructure based on the line skeleton, and performs global and local fusion to obtain a global-local combined skeleton model; Multi-level semantic constraint rule construction module: Constructs multi-level semantic constraint rules based on a global-local skeleton model; the specific implementation steps are as follows: Step 5.1: Construct spatial layout semantic constraint rules. Specifically, first define spatial layout constraint rules to constrain the horizontal layout of the BIM model of the component along the centerline of the route. Then, use the XML Schema specification language to standardize the spatial layout constraint rules to form spatial layout semantic constraint rules. Finally, use the spatial layout semantic constraint rules to constrain the horizontal layout of the BIM model of the component. Step 5.2: Construct spatial topological semantic constraint rules. Specifically, first, define spatial topological constraint rules to constrain the combination, connection, and vertical distribution relationships of components within the BIM model of the component. Then, use the XMLSchema specification language to standardize and express the spatial topological constraint rules to form spatial topological semantic constraint rules. Finally, use the spatial topological semantic constraint rules to constrain the combination, connection, and vertical distribution relationships within the BIM model of the component, so that when the BIM model of each component is mapped, the internal topological relationships of each component's BIM model conform to the correct topological relationships, achieving continuity in the vertical direction. Step 5.3: Construct spatial attitude semantic constraint rules. Specifically, first define spatial attitude constraint rules to constrain the attitude of the BIM model of the component itself, then use XML Schema to standardize the spatial attitude constraint rules to form spatial attitude semantic constraint rules, and finally use the spatial attitude semantic constraint rules to constrain the position and attitude of the component model itself, so that when the BIM model of each component is mapped, the BIM model of each component has the correct attitude, thus achieving accurate mapping. GIS Scene Mapping and Update Module: Based on the modified IFC component library, OBJ component library, information library, node mapping between the OBJ component library and the modified IFC component library and information library, global-local combined skeleton model, and multi-level semantic constraint rules, the BIM model is mapped to the GIS scene, and finally the GIS scene is obtained and updated in real time.

Images