A mapping association method and system for 3D tile and BIM vector model editing

Abstract

The application discloses a kind of 3D tile and BIM vector model editing mapping association method and system, by fusing the vector geometry information and topological relation of building information model to generate grid model, the efficient grid expression of vector data is realized;Subsequently, the node structure of grid model is parsed, attribute data is extracted and stored into database, a node attribute set is formed, and a unique identifier is generated for each node, a mapping relationship table is established to clearly record the correlation between nodes;On this basis, according to the node correlation set, the basic component unit is extracted from the grid model, and the customized three-dimensional tile is generated by combining attribute mapping, and the tile bottom binding is completed.The application effectively realizes the high-precision bidirectional association and rapid local update between BIM vector data and three-dimensional tile, significantly improves the efficiency and accuracy of building information model in web large-scale visualization, interactive editing and data synchronization.

Description

Technical Field

[0001] This invention relates to the field of building information modeling technology, and in particular discloses a mapping and association method and system for editing 3D tiles and BIM vector models. Background Technology

[0002] Building Information Modeling (BIM) technology plays an increasingly important role in the engineering and construction field. It integrates the geometry, spatial relationships, and various attribute information of buildings to provide support throughout the entire lifecycle, including design, construction, and operation and maintenance. With the continuous emergence of large-scale and complex building projects, the demand for efficient display and real-time interaction of BIM models in 3D scenes is growing stronger, especially in large-scale urban or industrial scenarios where BIM models need to maintain high accuracy while achieving fast loading and browsing.

[0003] Current mainstream 3D visualization technologies heavily utilize 3D tile formats to process massive amounts of model data. This approach significantly improves rendering performance and user experience through layered and tile-based loading. However, in practical applications, 3D tiles are essentially an optimized rasterized representation that discretizes the original building component information into geometric tiles and textures, severely fragmenting the model's original vector structure and rich attribute information. While designers or engineers can see realistic building appearances when browsing 3D scenes, it's difficult to directly select a specific component and view or modify its detailed attributes such as material, specifications, and installation location. They also cannot perform precise editing adjustments like with traditional vector models. This information fragmentation transforms 3D scenes from display tools into mere "image viewing" tools, failing to support real-world business processes that require frequent attribute queries and modifications.

[0004] A deeper problem lies in the lack of a stable correspondence between 3D tiles and the original vector model. Suppose an operations engineer discovers an incorrect labeling of a pipe in a 3D scene and wants to immediately correct its diameter and material properties. Current technical approaches typically involve exiting the 3D viewer, returning to specialized BIM software to find the corresponding vector component, modifying it, then re-exporting the entire model, re-slicing it into tiles, and uploading again. This process is time-consuming, cumbersome, and prone to causing a surge in server load or prolonged scene unavailability due to full updates. If attempting to edit directly on the 3D tiles, the lack of a reliable mapping to vector nodes means any changes cannot be accurately reflected in the original attribute database, ultimately leading to data inconsistencies or even the loss of critical information.

[0005] This artificially fragmented business chain—from viewing and locating to editing and updating—has become a key issue hindering the real-world application of 3D BIM models in actual projects. Summary of the Invention

[0006] This invention provides a mapping and association method and system for editing 3D tiles and BIM vector models, aiming to solve at least one of the defects existing in the prior art.

[0007] One aspect of the present invention relates to a mapping and association method for editing 3D tiles and BIM vector models, comprising the following steps:

[0008] S100: Integrate the geometric information and topological relationships of building information model vector data to generate a raster model;

[0009] S200. Analyze the node structure of the raster model to extract attribute data, store the attribute data in the database, and determine the node attribute set;

[0010] S300. Generate unique identifiers for nodes in the node attribute set and establish a mapping relationship table. Obtain the node association relationship set based on the mapping relationship table.

[0011] S400: Extract basic component units from the grid model based on the set of node association relationships, and generate customized three-dimensional tiles by combining attribute mapping to complete the bottom layer binding of the tiles;

[0012] S500: Obtain node identifiers by reverse querying the underlying binding relationship of the tile bottom layer, and use the restoration algorithm to process the raster data to obtain the edited vector model;

[0013] S600. Based on the modified data of the edited vector model, locally regenerate the 3D tiles and update the attribute mapping relationship to obtain the updated mapping association system.

[0014] Furthermore, the building information model vector data includes the vector boundary data of the building information model components, and step S100 includes:

[0015] S110. Obtain the vector boundary data of the building information model components, parse the vertex coordinates of the vector boundaries, and establish a set of geometric feature descriptions.

[0016] S120. Based on the geometric feature description set, identify the topological adjacency and containment relationships between components, and construct a global topological relationship graph;

[0017] S130. Discretize the coverage area of ​​the global topological relationship map to generate an initial voxel mesh structure;

[0018] S140. If the voxel grid in the initial voxel grid structure intersects with the vector boundary, then map the attributes to the voxel grid.

[0019] S150 aggregates the mapped voxel mesh and outputs a raster model that integrates geometric information and topological relationships.

[0020] Further, step S200 includes:

[0021] S210. Obtain a grid model that integrates geometric information and topological relationships. Use an octree depth-first traversal algorithm to scan the grid model to extract the original attribute data sequence.

[0022] S220. Parse the original attribute data sequence and convert it into structured attribute records according to the field mapping rules;

[0023] S230. Write the structured attribute records into the database storage table to generate a persistent attribute data table.

[0024] S240. Traverse the field metadata of the persistent attribute data table, aggregate the non-null field names and data type definitions in the field metadata to determine the node attribute set.

[0025] Further, step S300 includes:

[0026] S310. Traverse the node attribute set to extract spatial location coordinates and attribute feature data, and use a hash algorithm to combine the spatial location coordinates and attribute feature data to generate a globally unique identifier.

[0027] S320. Construct a node index mapping table based on a globally unique identifier. The node index mapping table contains the correspondence between node index keys and original storage addresses.

[0028] S330. Retrieve the node index mapping table to obtain the index keys of the neighboring nodes of the current node, and parse the spatial topology connection rules between the neighboring nodes and the current node.

[0029] S340. If the spatial distance between the spatial location vector of a neighboring node and the spatial location vector of the current node is less than or equal to a preset distance threshold, and the connection relationship between the spatial location vector of a neighboring node and the spatial location vector of the current node satisfies a preset connectivity condition, then a node adjacency tuple is generated, and all node adjacency tuples are aggregated to obtain a set of node association relationships.

[0030] Further, step S400 includes:

[0031] S410. Generate raster voxel clusters by aggregating voxel units based on the node association set, and the node association set represents the voxel topological connection;

[0032] S420. Construct a component solid mesh for the raster voxel cluster, and map the attribute feature vectors to the component solid mesh to generate a component object with attributes;

[0033] S430. Construct a spatial index tree based on the component object with attributes, traverse the spatial index tree to generate tile load data and establish a coordinate transformation interface to complete the bottom-level binding of the customized 3D tile.

[0034] Further, step S500 includes:

[0035] S510: Locate customized 3D tiles based on interactive commands, and extract unique node identifiers from the underlying binding relationships of the tile's bottom layer;

[0036] S520. Obtain the raster voxel cluster and attribute feature vector containing voxel topology connection information through the unique node identifier;

[0037] S530: Analyze the raster voxel clusters and combine them with attribute feature vectors to determine the boundary constraints. Input the shape restoration algorithm to fit and obtain the vector vertex coordinate sequence.

[0038] S540. Construct a geometric topological surface based on the fitted vector vertex coordinate sequence and map attribute feature vectors to generate the edited vector model.

[0039] Further, step S600 includes:

[0040] S610. Obtain the edited vector model and calculate the geometric bounding box. Traverse the preset three-dimensional tile spatial index tree to retrieve the tile encoding sequence to be updated that has spatial intersection with the geometric bounding box.

[0041] S620. Extract the corresponding original tile load based on the tile encoding sequence to be updated, and calculate the geometric topological difference between the edited vector model and the original tile load to determine the local reconstruction area;

[0042] S630: The discretized voxel algorithm is used to convert the vector geometric data in the local reconstruction area into binary stream data blocks that conform to the current tile level resolution. The binary stream data blocks are used to replace the original data segments corresponding to the tile encoding sequence to be updated, while retaining the data structure of the unchanged area to generate the reconstructed three-dimensional tile file.

[0043] S640. Extract the attribute feature vectors of the edited vector model, establish a new key-value pair mapping between the attribute feature vectors and the voxel indexes in the reconstructed 3D tile file through the unique node identifier, write the new key-value pair mapping into the global attribute association table, complete the synchronous update of the attribute mapping relationship, and output the updated mapping association system containing the locally regenerated tiles.

[0044] Another aspect of the present invention relates to a mapping and association system for editing 3D tiles and BIM vector models, used to perform the above-described mapping and association method for editing 3D tiles and BIM vector models, comprising:

[0045] The raster model generation module is used to integrate the geometric information and topological relationships of the building information model vector data to generate a raster model;

[0046] The node attribute set determination module is used to parse the node structure of the raster model to extract attribute data, store the attribute data in the database, and determine the node attribute set.

[0047] The node association set acquisition module is used to generate unique identifiers for nodes in the node attribute set and establish a mapping relationship table, and then obtain the node association set based on the mapping relationship table.

[0048] The tile bottom layer binding module is used to extract basic component units from the raster model based on the set of node association relationships, and combine them with attribute mapping to generate customized 3D tiles to complete the tile bottom layer binding.

[0049] The vector model acquisition module is used to obtain node identifiers through reverse lookup by binding the bottom layer of tiles, and to process raster data using a restoration algorithm to obtain the edited vector model.

[0050] The attribute mapping update module is used to locally regenerate 3D tiles and update attribute mapping relationships based on the modified data of the edited vector model, thereby obtaining an updated mapping association system.

[0051] The beneficial effects achieved by this invention are as follows:

[0052] This invention provides a mapping and association method and system for editing 3D tiles and BIM vector models. By integrating the vector geometric information and topological relationships of a Building Information Model (BIM) to generate a raster model, it achieves efficient rasterized representation of vector data. Subsequently, it analyzes the node structure of the raster model, extracts attribute data, and stores it in a database, forming a node attribute set. A unique identifier is generated for each node, and a mapping relationship table is established to clearly record the relationships between nodes. Based on this, basic component units are extracted from the raster model according to the node relationship set. Customized 3D tiles are generated by combining attribute mapping and the underlying layer of the tiles is then bound. This solves the business challenges of accurately matching traditional 3D tiles with original BIM attribute information and the difficulty in reverse synchronization after editing. In practical applications, users can directly edit the tiles, reverse-query node identifiers through the underlying layer binding, and use a restoration algorithm to convert the edited raster data into a modified vector model. Then, based on the modifications, the 3D tiles are locally regenerated and the attribute mapping relationships are updated, ultimately forming an efficient closed-loop mapping and association system. This invention effectively achieves high-precision bidirectional association and rapid local updates between BIM vector data and 3D tiles, significantly improving the efficiency and accuracy of large-scale visualization, interactive editing, and data synchronization of building information models on the Web. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating an embodiment of the mapping and association method for editing 3D tiles and BIM vector models according to the present invention.

[0054] Figure 2 This is a functional block diagram of an embodiment of the mapping and association system for editing 3D tiles and BIM vector models according to the present invention.

[0055] Explanation of icon numbers:

[0056] 10. Grid model generation module; 20. Node attribute set determination module; 30. Node association relationship set acquisition module; 40. Tile bottom layer binding completion module; 50. Vector model acquisition module; 60. Attribute mapping relationship update module. Detailed Implementation

[0057] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0058] like Figure 1 As shown, the first embodiment of the present invention proposes a mapping and association method for editing 3D tiles and BIM vector models, including the following steps:

[0059] Step S100: Integrate the geometric information and topological relationships of the building information model vector data to generate a raster model.

[0060] As a core step in the mapping and association method between 3D tiles and BIM vector models, this step uses native BIM vector data as input and focuses on fusing geometric and topological information and converting vector to raster data. First, it extracts the geometric information (such as vertex coordinates, dimensions, shape, and spatial location) and topological relationships (such as adjacency, containment, and dependency relationships between components) of the components from the BIM vector data. Through algorithms such as vector discretization and voxel mesh mapping, the continuous BIM vector space is transformed into a raster structure with voxel meshes as the basic units. Geometric features and topological association information at corresponding locations are embedded in each raster unit, ultimately generating a raster model that integrates geometric information and topological relationships. This step solves the technical pain point of BIM vector models being difficult to directly adapt to 3D tile segmentation and rendering, establishing a data foundation for subsequent mapping and association.

[0061] Step S200: Analyze the node structure of the raster model to extract attribute data, store the attribute data in the database, and determine the node attribute set.

[0062] As the attribute data structuring step in this method, this step performs a refined analysis of the raster model generated in step S100. The core focus is on the node structure of the raster model (including vertices, feature points, and component-related nodes of the voxel mesh). Full-dimensional attribute data is extracted from each node—covering geometric attributes (node ​​coordinates, elevation, and size of the raster unit), topological attributes (BIM components corresponding to the node, and relationships between adjacent nodes), and business attributes (BIM native attributes such as component material, specifications, cost, and construction stage). The extracted attribute data is cleaned, standardized, and stored in a structured database (such as a spatiotemporal database or a relational database). All attribute information is then integrated according to the node dimension to form a node attribute set containing node identifiers, multi-dimensional attributes, and topological relationships. This step transforms the raster model from "spatial morphological data" to "structured attribute data," providing a reliable attribute basis for establishing subsequent mapping relationships.

[0063] Step S300: Generate unique identifiers for nodes in the node attribute set and establish a mapping relationship table, and obtain the node association relationship set based on the mapping relationship table.

[0064] As the core link in this method, this step is crucial for achieving bidirectional mapping between 3D tiles and the BIM vector model. First, a globally unique identifier (UUID) is assigned to each node in the node attribute set generated in step S200, ensuring that each node has a unique identity throughout the entire model system. Based on the node's unique identifier, the correspondence between nodes, between nodes and BIM components, and between nodes and grid units is analyzed, constructing a structured mapping relationship table (containing core fields such as source node ID, associated node ID, belonging component ID, grid unit number, and attribute mapping path). By traversing the mapping relationship table, all association logic such as hierarchy, dependency, and connection between nodes is derived and integrated, forming a node association relationship set (including subsets such as adjacency, containment, and dependency relationships). This step establishes a precise association link between grid nodes, BIM components, and tile units through the unique identifier and the mapping relationship table.

[0065] Step S400: Extract basic component units from the raster model based on the node association set, and generate customized three-dimensional tiles by combining attribute mapping to complete the bottom layer binding of the tiles.

[0066] As the tile customization and association binding step in this method, this core step completes the transformation and underlying association of "grid model → 3D tile": First, based on the node association relationship set in step S300, basic component units with independent engineering significance (such as a single beam, a single column, a wall, etc., corresponding to a set of associated nodes) are aggregated and extracted from the grid model; based on the business attributes and visualization requirements of BIM components, the basic component units are divided into blocks and hierarchical divisions to generate customized 3D tiles that conform to the 3D tile technical specifications (the tile data includes the gridded geometric information of the component units and node attribute mapping pointers); the key is to complete the underlying binding of the tiles - to deeply bind the block units of each 3D tile with the unique identifier and mapping relationship table of the node attribute set, so that the 3D tile has "visual form + attribute traceability capability", and can be directly associated with the corresponding BIM vector nodes, components and full attribute data through the tile.

[0067] Step S500: Obtain the node identifier by reverse querying the underlying binding relationship of the tile bottom layer, and use the restoration algorithm to process the raster data to obtain the edited vector model.

[0068] As the core of this method's reverse mapping process, this step solves the problem of disconnect between traditional 3D tile editing and BIM vector models: When a user performs visual editing on a 3D tile (such as modifying component size, moving position, or replacing material), the unique identifier of the node bound to the tile unit corresponding to the editing operation is obtained through the tile's underlying binding relationship completed in step S400. Using this node identifier as an index, the corresponding node attribute set and raster model data are retrieved from the database. Using raster-vector inverse transformation, topology reconstruction, and other restoration algorithms, the edited raster data is restored to a native vector format conforming to BIM standards. The restored vector data undergoes topology verification and attribute updates, ultimately generating the edited vector model, thus achieving a reverse mapping of "visual tile editing → synchronous update of BIM vector model".

[0069] Step S600: Based on the modified data of the edited vector model, locally regenerate the 3D tiles and update the attribute mapping relationship to obtain the updated mapping association system.

[0070] As a closed-loop optimization and system update step in this method, the core of this step is to achieve dynamic consistency maintenance of the mapping association system: First, the modified data of the vector model edited in step S500 (such as component size change values, attribute modification content, and node position adjustment information) is extracted. The "partial update" strategy is adopted to regenerate only the 3D tiles corresponding to the modified area (instead of reconstructing all tiles), which greatly reduces the computational cost. Simultaneously, the node association relationship and attribute mapping path in the mapping relationship table are adjusted according to the modified data, and the business attribute data in the node attribute set is updated. Finally, the locally regenerated 3D tiles, the updated mapping relationship table, and the node attribute set are integrated to form the updated mapping association system, ensuring that the 3D tiles and the BIM vector model always maintain geometric, attribute, and topological consistency after editing, completing the entire closed-loop process of "edit-restore-update".

[0071] Furthermore, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment, wherein the building information model vector data includes the vector boundary data of building information model components, includes step S100 as follows:

[0072] Step S110: Obtain the vector boundary data of the building information model components, parse the vertex coordinates of the vector boundaries, and establish a set of geometric feature descriptions.

[0073] The set of geometric feature descriptions is derived using the following formula:

[0074] (1);

[0075] In formula (1), Represents a set of geometric feature descriptions. Indicates the first Geometric feature description, This represents the total number of geometric features. The control logic of formula (1) is...

[0076] When processing Building Information Modeling (BIM), the first step is to obtain the vector boundary data of the components. This involves extracting the boundary contours of components such as walls, beams, and columns from BIM software. For example, the boundary data of the walls of a rectangular room includes the coordinates of four vertices, such as (0, 0, 0), (5, 0, 0), (5, 4, 0), and (0, 4, 0). By parsing these vertex coordinates, a set of geometric feature descriptions is established, such as calculating the length, angle, and curvature of each boundary, forming a set describing the geometry of the wall, thus providing basic data support for subsequent topology analysis. This method ensures the accuracy of the data and effectively captures the spatial characteristics of components in actual building simulations, avoiding calculation errors in later stages.

[0077] Step S120: Based on the geometric feature description set, identify the topological adjacency and containment relationships between components and construct a global topological relationship graph.

[0078] The global topological relationship graph is derived using the following formula:

[0079] (2);

[0080] In formula (2), This represents the final global topology graph. Geometric features describe the size of the set. Represents a single geometric feature description within a set. This represents the local topological relationship map matrix constructed based on the geometric feature description. The control logic of formula (2) is to traverse and aggregate all geometric features parsed from the vector boundary of BIM components to form a complete set of geometric feature descriptions, which provides an accurate geometric data foundation for the mapping and association between 3D tiles and BIM vector models. As the core data preprocessing module of the BIM and 3D tile fusion system, it provides reliable data support for efficient editing, updating and visualization.

[0081] Based on the aforementioned set of geometric features, topological adjacency and containment relationships between components are identified. For example, suppose there are two adjacent wall components, one of which shares a boundary with the other. By comparing their coordinate sets, it can be determined whether they intersect or are adjacent. If a component, such as a column inside a room, is completely contained within the boundary of another component, such as a floor, it is marked as containing. A global topological graph is then constructed. This graph structure represents components, and edges indicate adjacency or containment. For example, wall nodes connect to adjacent door and window nodes, forming a global topological graph. This global topological graph helps in understanding the overall connectivity of a building, and in urban planning, it can optimize spatial layout design and improve the structural stability of buildings.

[0082] Step S130: Discretize the coverage area of ​​the global topological relationship map to generate an initial voxel mesh structure.

[0083] The initial voxel mesh structure is generated using the following formula:

[0084] (3);

[0085] In formula (3), This represents the initial voxel grid structure. express Direction index, express Direction index, express Direction index, express Number of directional grids, express Number of directional grids, express Number of directional grids. The control logic of formula (3) is to traverse and aggregate all geometric features parsed from the vector boundary of BIM components to form a complete set of geometric feature descriptions, which provides a topological basis for the mapping from BIM editing to 3D tile updates, and realizes accurate incremental updates. As the core module of the method of mapping and associating 3D tiles with BIM vector models, it provides topological support for efficient and reliable scene editing and updating.

[0086] When discretizing the area covered by the global topological map, the entire building space is divided into a uniform voxel grid. For example, a 10-meter-high floor area is divided into a cubic grid with 1-meter sides, generating thousands of voxels to form the initial voxel grid structure. This discretization is similar to transforming continuous space into discrete units, which is convenient for computationally intensive tasks and can quickly assess the stress distribution of the structure in earthquake simulation.

[0087] Step S140: If the voxel grid in the initial voxel grid structure intersects with the vector boundary, then map the attributes to the voxel grid.

[0088] The following formula is used to define the intersection indicator function between the voxel mesh and the vector boundary:

[0089] (4);

[0090] In formula (4), This indicates the function representing the intersection of the voxel mesh and the vector boundary. This represents the voxel grid in the initial voxel grid structure. Represents the vector boundary, if the voxel mesh in the initial voxel mesh structure With vector boundary If they intersect, the value is 1; otherwise, it is 0. This is used to determine whether attribute mapping should be performed. The control logic of formula (4) is to determine whether the voxel mesh intersects with the vector boundary and use this as the execution condition for attribute mapping. Only voxels that intersect with the vector boundary are subject to attribute mapping, avoiding unnecessary calculations and ensuring that the attributes of the voxelized 3D tile data are consistent with those of the original BIM vector boundary, thus providing accurate attribute information for subsequent 3D rendering and analysis.

[0091] If the initial voxel mesh intersects with the vector boundary, the properties are mapped to the voxel mesh. For example, when a voxel partially covers a wall boundary, the wall's material properties, such as concrete density, are mapped onto that voxel mesh. By calculating the proportion of the intersection volume, attribute values ​​are assigned to ensure that the voxel mesh carries accurate geometric information.

[0092] Step S150: Aggregate the mapped voxel mesh and output a raster model that integrates geometric information and topological relationships.

[0093] The raster model is derived using the following formula:

[0094] (5);

[0095] In formula (5), Represents a raster model. Indicates the output function. The voxel mesh represents the completed mapping. The control logic of formula (5) is to aggregate and encode the voxel mesh that has completed attribute mapping, and output a raster model that integrates geometric and topological information. Through encoding and format conversion, the storage volume and transmission efficiency of the data are optimized. As the final output of the method for editing and mapping association between 3D tiles and BIM vector models, it provides a reliable data foundation for subsequent scene applications and updates.

[0096] After aggregating the mapped voxel mesh, a raster model that integrates geometric information and topological relationships is output. This involves merging adjacent voxels to form larger raster cells while preserving topological relationships, such as adjacent voxels sharing boundary properties. In one implementation, this raster model is used in virtual reality building walkthroughs, enabling real-time rendering of complex scenes, improving interaction efficiency, and providing a data foundation for subsequent thermodynamic analysis, thus achieving the technical effect of optimizing building energy consumption.

[0097] Furthermore, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment includes step S200:

[0098] Step S210: Obtain the raster model that integrates geometric information and topological relationships, and use the octree depth-first traversal algorithm to scan the raster model to extract the original attribute data sequence.

[0099] The original attribute data sequence is obtained using the following formula:

[0100] (6);

[0101] In formula (6), Represents the original attribute data sequence. Represents raster model attributes, This indicates the traversal order of the octree. The extraction function is represented by formula (6). The control logic of formula (6) is to rearrange the three-dimensional attribute information in the raster model into a one-dimensional linear sequence through the depth-first traversal of the octree. The fixed octree traversal order ensures the repeatability and comparability of the attribute sequence. As a key data preprocessing step in the method of editing and mapping association between 3D tiles and BIM vector models, it provides a foundation for efficient incremental updates and data management.

[0102] The acquired raster model, which integrates geometric information and topological relationships, is essentially a three-dimensional attribute field. For example, in a building structural health monitoring scenario, this raster model consists of millions of voxels. Each voxel not only contains its spatial location but also incorporates data such as material properties, component types, and stress states inherited from the original building information model. To efficiently access and extract this massive amount of data, an octree depth-first traversal algorithm is used for scanning. Specifically, an octree is a spatial partitioning data structure that recursively divides the entire three-dimensional raster space into eight sub-cubes. Depth-first traversal means that the octree depth-first traversal algorithm will delve into a branch all the way to the lowest leaf node voxel, and then backtrack to traverse other branches. For example, the octree depth-first traversal algorithm starts from the root node representing the entire building model and determines whether the node is composed entirely of voxels with the same attribute. If not, it is divided into eight child nodes, and the same judgment and division are continued for the first child node until an indivisible single voxel node is reached. During this process, whenever a leaf node voxel is visited, its spatial coordinates and all attribute values ​​are extracted in a predetermined order, forming a linear sequence of raw attribute data. This process ensures the integrity of data extraction and the orderliness of spatial traversal.

[0103] Step S220: Parse the original attribute data sequence and convert it into structured attribute records according to the field mapping rules.

[0104] Structured attribute records are derived using the following formula:

[0105] (7);

[0106] In formula (7), Represents structured attribute records, Represents the original attribute data sequence. Indicates the field mapping rules. This represents the conversion function that parses and follows the mapping rules. The control logic of formula (7) is to parse the original attribute data sequence and convert it into a structured attribute record with clear semantics according to the field mapping rules. This converts the unstructured original data sequence into a structured record with clear semantics, improving the readability and usability of the data. As a key data processing step in the method of editing and mapping association between 3D tiles and BIM vector models, it provides a reliable foundation for efficient data management and application.

[0107] Next, the parsed raw attribute data sequence needs to be structured. The raw attribute data sequence is typically a byte stream arranged in a fixed format, containing multiple fields such as voxel coordinates, material number, density value, and temperature value, but these fields are mixed together. According to predefined field mapping rules, these raw attribute data sequences are parsed into independent fields with clear meaning. For example, the mapping rule stipulates that the first 12 bytes of the sequence represent the center coordinates of the voxel, the next 4 bytes represent the component type code, and the following 8 bytes represent the currently calculated stress value. Through parsing, a raw byte block is transformed into a clear structured attribute record such as "coordinates, component type, stress value".

[0108] Step S230: Write the structured attribute records into the database storage table to generate a persistent attribute data table.

[0109] The persistent property data table is derived using the following formula:

[0110] (8);

[0111] In formula (8), Represents a persistent attribute data table. This represents a persistent function. Represents a database storage table. Represents structured attribute records, Given conditions, this formula is used to perform persistence operations on the database storage table based on structured attribute records, generating a persistent attribute data table. The control logic of formula (8) is to perform persistence operations on the database storage table based on structured attribute records, generating a persistent attribute data table. Through transactions, the consistency between BIM attribute data and database storage is guaranteed, providing a reliable data source for subsequent applications. As a key data management step in the method of editing and mapping association between 3D tiles and BIM vector models, it provides a reliable foundation for efficient data query, update and system integration.

[0112] These records are then written in batches to a specific storage table in the database, such as a table named "Voxel Attribute Snapshot," thus achieving persistent data storage and laying the foundation for subsequent queries and analysis. To dynamically construct data queries or visualization interfaces, it is necessary to automatically identify available data fields from the generated persistent attribute data table. This process is achieved by traversing the field metadata of the persistent attribute data table. The field metadata describes the name, data type, and constraints of each column in the persistent attribute data table.

[0113] Step S240: Traverse the field metadata of the persistent attribute data table, aggregate the non-null field names and data type definitions in the field metadata to determine the node attribute set.

[0114] The set of node attributes is determined using the following formula:

[0115] (9);

[0116] In formula (9), Represents a collection of node attributes. Represents a persistent attribute data table. Represents field metadata. This indicates a traversal operation. Represents a field. Indicates the field name. This indicates the data type definition. This indicates the filtering condition, retaining only fields whose names are not empty and excluding invalid or placeholder fields. This formula is used to determine the node attribute set by traversing the field metadata and aggregating the names and data types of non-empty fields. The control logic of formula (9) is to traverse the field metadata of the persistent attribute data table, filter out invalid fields, aggregate the names and types of valid fields, and form a node attribute set. By traversing and filtering the metadata, the attribute set is automatically generated, reducing the workload of manual configuration and the probability of errors. As a key step in the method of editing and mapping association between 3D tiles and BIM vector models, it transforms relational data into graph data and provides a reliable attribute definition foundation for subsequent knowledge graph applications.

[0117] The mapping and association system scans the field metadata information of the persistent attribute data table, filters out all non-empty fields, and collects their names and corresponding data type definitions. From the "Voxel Attribute Snapshot" table, it aggregates a series of field definitions such as "Center Point X Coordinate, Floating-point Number", "Center Point Y Coordinate, Floating-point Number", "Material Name, String", and "Young's Modulus, Floating-point Number". This set is called the node attribute set, which clearly defines the attribute dimensions of each node in the current data model. This node attribute set is crucial for subsequent data binding and dynamic analysis. For example, in structural simulation software, engineers can flexibly choose whether to map the "Stress Value" field or the "Displacement" field to the color rendering channel of the 3D model based on this node attribute set, thereby intuitively observing different physical field distributions.

[0118] Preferably, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment includes step S300:

[0119] Step S310: Traverse the node attribute set to extract spatial location coordinates and attribute feature data, and use a hash algorithm to combine the spatial location coordinates and attribute feature data to generate a globally unique identifier.

[0120] Globally unique identifiers are derived using the following formula:

[0121] (10);

[0122] In formula (10), Represents a globally unique identifier. Represents a hash algorithm. The combination represents the spatial location coordinates and attribute feature data. The control logic of formula (10) is to combine the spatial location and attribute feature data and generate a globally unique identifier through a hash algorithm. By combining the spatial location and attribute features through hashing, the uniqueness of the identifier in the entire system is ensured. As the core identification mechanism of the 3D tile and BIM vector model editing mapping association method, it provides a reliable technical foundation for accurate incremental updates and data traceability.

[0123] Combined spatial location coordinates and attribute feature data This can be derived from the following formula:

[0124] (11);

[0125] In formula (11), This represents the spatial location coordinates and attribute feature data of the combination. Represents spatial location coordinates, This indicates a combination and splicing operation. The control logic of formula (11) is to concatenate spatial location coordinates and attribute feature data according to fixed rules to generate composite data for hash calculation. This concatenation of spatial location coordinates and attribute feature data according to fixed rules to generate composite data for hash calculation serves as a prerequisite for generating a globally unique identifier, providing a reliable data foundation for the accurate mapping of 3D tiles and BIM vector model editing.

[0126] Spatial location coordinates and attribute feature data This can be derived from the following formula:

[0127] (12);

[0128] In formula (12), Represents spatial location coordinates, Represents attribute feature data, Represents a node The set of attributes This indicates the traversal extraction operation. The control logic of formula (12) is to traverse the set of node attributes, separate and extract the spatial location coordinates and attribute feature data, and separate spatial information from attribute information, so that subsequent processing (such as hashing and querying) is more targeted and efficient. As a key pre-step for generating a globally unique identifier, it provides a reliable data foundation for the accurate mapping of 3D tiles and BIM vector model editing.

[0129] The node attribute set originates from a 3D digital model of an urban underground pipe network. Each node in this 3D digital model represents a pipe segment connection point or valve, and its attribute set includes multiple fields such as "geographical longitude," "geographical latitude," "burial depth and elevation," "pipe diameter," "pipe material," and "fluid pressure." To build an efficient index, it is necessary to extract the "longitude," "latitude," and "elevation" data that constitute the spatial location coordinates from these fields, while selecting "pipe diameter" and "pipe material" as key attribute feature data. Specifically, the mapping and association system reads these field values ​​for each node and concatenates them into a data string in a predetermined order, such as "116.403, 39.924, 15.2, 300, cast iron." Subsequently, the SHA256 hash algorithm is applied to this concatenated string to generate a fixed-length, seemingly random hexadecimal string, which serves as the globally unique identifier for that node. This identifier is highly unique and deterministic. As long as the spatial location and key attributes of the node remain unchanged, its identifier will remain unchanged, which lays the foundation for subsequent rapid retrieval and comparison.

[0130] Step S320: Construct a node index mapping table based on the globally unique identifier. The node index mapping table contains the correspondence between node index keys and original storage addresses.

[0131] The node index mapping table is derived using the following formula:

[0132] (13);

[0133] In formula (13), Represents the node index mapping table. Indicates the index key of the current node. The formula (13) represents the original storage address and is used to establish the correspondence between the node index key and the original storage address. The control logic of the formula (13) is to establish a one-to-one correspondence between the node index key and the original storage address, thereby constructing an efficient index mapping table, converting the direct full table scan into an index lookup with O(1) complexity, which significantly improves the access performance in big data scenarios; the index layer isolates the upper-layer application from the underlying storage details, making the system more scalable and maintainable; as the core index mechanism of the 3D tile and BIM vector model editing mapping association method, it provides a reliable technical foundation for efficient data management and accurate incremental updates.

[0134] Current node index key This can be derived from the following formula:

[0135] (14);

[0136] In formula (14), Indicates the index key of the current node. Represents a hash function. The formula (14) represents a globally unique identifier (UID), which is used to generate node index keys based on the UID. The control logic of formula (14) is to perform a secondary hash or transformation on the globally unique identifier (UID) to generate node index keys that adapt to the index structure. Through hash transformation, the globally unique identifier is adapted to the key value requirements of the index system, improving the compatibility and performance of the index. Through compression and uniform distribution, the index storage overhead is reduced, and the query efficiency is improved. As a prerequisite for the construction of the node index mapping table, it provides a reliable technical foundation for the accurate mapping and efficient data management of 3D tiles and BIM vector model editing.

[0137] Original storage address This can be derived from the following formula:

[0138] (15);

[0139] In formula (15), This represents the address extraction function. Indicates the original storage address. The UID represents a globally unique identifier used to obtain the original storage address of the node. The control logic of formula (15) performs a secondary hash or transformation on the globally unique identifier (UID) to generate a node index key that adapts to the index structure. The data storage location is quickly located through the UID, avoiding a full table scan and improving data access efficiency. The address extraction function isolates the upper-layer application from the underlying storage details, making the system more scalable and maintainable. As a key step in the construction of the node index mapping table, it provides a reliable technical foundation for the accurate mapping and efficient data management of 3D tiles and BIM vector model editing.

[0140] Based on the generated globally unique identifiers, a node index mapping table is constructed. This node index mapping table is essentially a key-value database, where the key is the hash identifier mentioned above, and the value is the physical storage address of the corresponding node in the original database or file system. For example, the identifier "a1b2c3d4..." maps to database record ID "102457", or to offset "byte 2048" in the file "pipe_data.bin". This node index mapping table allows the system to quickly locate the complete original data of a node based on its essential characteristics, without performing complex spatial range queries.

[0141] Step S330: Retrieve the node index mapping table to obtain the index keys of the neighboring nodes of the current node, and parse the spatial topology connection rules between the neighboring nodes and the current node.

[0142] The neighbor node index key is derived using the following formula:

[0143] (16);

[0144] In formula (16), Indicates the neighbor node index key. Indicates the retrieval function. Represents the node index mapping table. The formula (16) represents the index key of the current node. It is used to retrieve the index keys of the neighboring nodes of the current node by searching the node index mapping table. The control logic of formula (16) is to retrieve the index keys of the neighboring nodes of the current node by searching the node index mapping table and to parse their spatial topology connection rules. The neighboring nodes are quickly obtained through the index mapping table, which enables the system to have topology awareness and provides support for intelligent editing and impact analysis. The affected nodes are accurately located through topology relationships, avoiding full updates and significantly improving the efficiency of 3D tile updates. As the core topology analysis step of the 3D tile and BIM vector model editing mapping association method, it provides a reliable technical foundation for efficient and intelligent scene editing.

[0145] When connectivity analysis is required, such as analyzing the gas supply path of a gas pipeline network, the mapping and association system searches this index mapping table to find the neighbors of the current node. For example, for a valve node, the algorithm searches the node index mapping table for other node identifiers with the same or compatible "pipe diameter" and "pipe material" within a certain spatial tolerance range, based on its spatial coordinates and pipe diameter. These found identifiers are the potential neighbor node index keys. Next, the mapping and association system needs to parse the spatial topology connection rules between these neighbor nodes and the current node. In the pipeline model, connection rules include "direct docking on the same plane," "connection via tee fittings," or "vertical connection between manholes," etc. The mapping and association system loads the complete geometric data of the two nodes, calculates the spatial distance, angle, and interface type at the connection point, and compares it with predefined engineering specifications.

[0146] Step S340: If the spatial distance between the spatial location vector of a neighboring node and the spatial location vector of the current node is less than or equal to a preset distance threshold, and the connection relationship between the spatial location vector of a neighboring node and the spatial location vector of the current node satisfies a preset connectivity condition, then a node adjacency tuple is generated, and all node adjacency tuples are aggregated to obtain a set of node association relationships.

[0147] The node adjacency tuple is obtained using the following formula:

[0148] (17);

[0149] In formula (17), Represents the set of adjacency pairs of nodes. Indicates the index key of the current node. Indicates the neighbor node index key. This represents the spatial location vector of the current node. Represents the spatial location vector of neighboring nodes. Represents a spatial distance function. This indicates the preset distance threshold. The function represents connectivity judgment, and a return value of 1 indicates that the preset connectivity conditions are met. The control logic of formula (17) is based on spatial distance and connectivity conditions to filter out effective node adjacency pairs, construct the edge set of the topology graph, and ensure the accuracy of adjacency relationships and avoid misjudgment through dual conditions (spatial distance + connectivity); it provides topology awareness capability for BIM editing, realizes intelligent influence range analysis and incremental updates; as the core topology construction step of the 3D tile and BIM vector model editing mapping association method, it provides a solid technical foundation for efficient and reliable scene management.

[0150] The set of node relationships is derived using the following formula:

[0151] (18);

[0152] In formula (18), Represents the set of node relationships. Denotes the set of all adjacent pairs. Representing adjacency pairs, this formula aggregates adjacency pairs based on existence conditions, directly defining the set of node association relationships. The control logic of formula (18) is to directly aggregate adjacency pairs that meet the conditions into a set of node association relationships. By directly aggregating, it ensures that the set of association relationships is completely consistent with the set of adjacency pairs, avoiding deviations in the definition of topological relationships; it simplifies the conversion process from adjacency relationships to association relationships, improving system efficiency; as the final definition of topological relationships in the 3D tile and BIM vector model editing mapping association method, it provides a reliable foundation for knowledge graph construction and intelligent editing.

[0153] If the calculated spatial relationship meets the preset connectivity conditions, such as a distance of less than 1 cm and matching interface types, then the two nodes are considered topologically connected. At this point, the mapping and association system generates a node adjacency tuple, which is a combination of the globally unique identifiers of the two nodes, such as "<identifier A, identifier B>". By traversing the entire node index mapping table, the above neighbor retrieval, rule parsing, and connectivity judgment process is repeated for each node, ultimately aggregating the set of all connectivity relationships in the entire network system, i.e., the complete set of node association relationships. This set of relationships is the core data foundation for advanced applications such as network traffic analysis and fault impact range simulation.

[0154] Furthermore, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment includes step S400 as follows:

[0155] Step S410: Generate raster voxel clusters by aggregating voxel units based on the node association relationship set. The node association relationship set represents the voxel topological connection.

[0156] Raster voxel clusters are derived using the following formula:

[0157] (19);

[0158] In formula (19), represents a grid voxel cluster, Represents the set of node relationships. Represents rasterization mapping, The formula (19) represents a voxel unit and is used to generate raster voxel clusters. The control logic of formula (19) is to aggregate topologically connected voxel units into continuous raster voxel clusters based on the set of node association relationships. Through voxel clusters, more fine-grained incremental updates are achieved, avoiding full tile redrawing. As the core spatial aggregation step of the 3D tile and BIM vector model editing mapping association method, it provides a solid technical foundation for efficient and intelligent scene management and updates.

[0159] The process of aggregating voxel units based on node association sets is based on understanding structural components in a building information model (BIM). For example, the geometry of a digital model of a reinforced concrete column is discretized into thousands of tiny cubes, i.e., voxel units. These voxel units themselves only contain basic information such as spatial location and material type. The mapping association system first uses the set of node associations generated in previous steps, which characterizes the topological connectivity between voxel units, such as which voxels belong to the same reinforcing bar and which voxels belong to the concrete surrounding the reinforcing bar. The mapping association system scans all voxels, merging voxel units with connectivity and consistent attributes together to form a continuous raster voxel cluster representing a single physical component. Specifically, for a beam component, the algorithm identifies all spatially adjacent voxels labeled "beam concrete C30" and aggregates them from a scattered point cloud state into a complete beam-shaped voxel block.

[0160] Step S420: Construct a component entity mesh for the raster voxel cluster, and map the attribute feature vectors to the component entity mesh to generate a component object with attributes.

[0161] Component objects with attributes are derived using the following formula:

[0162] (20);

[0163] In formula (20), Represents a component object with attributes. Represents the attribute feature vector. Represents the component solid mesh, The formula (20) represents the mapping function, which describes the attribute generation process of mapping feature vectors to the mesh. The control logic of formula (20) is to map the attribute feature vectors to the component entity mesh, generate an attributed component object that integrates geometry and attributes, realize the deep binding of BIM component geometric information and attribute information, and ensure the consistency and integrity of the data; it enables the rasterized 3D tile data to have the semantic attributes of BIM, improves the usability and analyzability of the data; as the final output of the 3D tile and BIM vector model editing mapping association method, it provides a reliable technical foundation for efficient BIM editing, updating and application.

[0164] Component solid mesh This can be derived from the following formula:

[0165] (twenty one);

[0166] In formula (21), Represents the component solid mesh, represents a grid voxel cluster, The formula (21) represents the construction function, which describes the process of meshing voxel clusters. The control logic of formula (21) is to perform mesh reconstruction on the raster voxel clusters to generate continuous component solid meshes, converting discrete voxel data into continuous polygon meshes, which significantly improves the accuracy and smoothness of geometric representation. As a key conversion step from raster voxels to BIM component geometric representation, it provides a reliable geometric basis for the accurate mapping of 3D tiles and BIM vector model editing.

[0167] Subsequently, the surface geometry of the aggregated raster voxel clusters needs to be constructed, i.e., the component solid mesh. This process involves extracting triangular facets from the voxel boundaries. For example, the mapping and association system traverses the outer voxels representing the aforementioned reinforced concrete column, checking the six faces of each voxel cube and determining whether there are voxels with different attributes or empty faces in adjacent positions. If a face is empty or has other components on its outer side, then that face is determined to be the surface of the component. Next, these exposed faces are transformed into fine triangular meshes, thereby generating a lightweight column mesh model that can be used for 3D rendering. While generating the triangular mesh, attribute feature vectors need to be mapped to the triangular mesh. The attribute feature vectors contain information such as "design strength grade," "production batch number," "construction date," and "fire resistance rating" extracted from the original data. The mapping and association system creates a data structure for this newly generated column mesh object, binding the geometric mesh to these attribute fields, thereby generating a component object with attributes, which simultaneously possesses a visual geometric shape and queryable business attributes.

[0168] Step S430: Construct a spatial index tree based on the component object with attributes, traverse the spatial index tree to generate tile load data and establish a coordinate transformation interface to complete the bottom layer binding of the customized 3D tile.

[0169] The underlying tile bonding is determined by the following formula:

[0170] (twenty two);

[0171] In formula (22), This indicates that the bottom layer of the tile is bonded. Indicates the binding function. This indicates customized 3D tiles. This indicates the tile load data. This indicates the coordinate transformation interface. Represents the original coordinates. The coordinates are represented after transformation. The control logic of formula (22) is to associate and bind the customized 3D tile, its load data and coordinate transformation interface to form a tile bottom layer structure that can be used directly. It separates the structural definition, content data and spatial coordinates of the tile, which improves the flexibility and maintainability of the system. Through the coordinate transformation interface, BIM tiles can be seamlessly adapted to application scenarios of different spatial reference systems. As the final implementation step of the 3D tile and BIM vector model editing mapping association method, it provides complete technical support for efficient and reliable BIM tile application.

[0172] Tile load data This can be derived from the following formula:

[0173] (twenty three);

[0174] In formula (23), Indicates tiles Load data, This represents the traversal function. Represents a spatial index tree, Indicates tiles The boundary area Represents a node The load value, This represents the total load of nodes within the boundary of the traversal tree. The control logic of formula (23) is to traverse all nodes within the tile boundary in the spatial index tree, aggregate their load values, and generate the total load data of the tile. Through the traversal of the spatial index tree, the tile data is quickly and accurately aggregated, avoiding full data scanning. As the core data source for the bottom layer binding of the tile, it provides a reliable data foundation for the accurate mapping and efficient application of 3D tiles and BIM vector model editing.

[0175] Spatial Index Tree This can be derived from the following formula:

[0176] (twenty four);

[0177] In formula (24), Represents a spatial index tree, Indicates the constructor function. Indicates the first The position and properties of a component object with attributes. This represents the total number of component objects. The control logic of formula (24) is based on all component objects with attributes. It constructs an efficient spatial index tree through a spatial partitioning algorithm, reducing the query complexity of spatial data from linear scanning. Reduced to logarithmic level It significantly improves performance in big data scenarios; through hierarchical spatial division, it achieves orderly management of complex BIM scenarios, providing a foundation for subsequent tile generation and updates; as the core pre-structure for tile load data generation and tile bottom layer binding, it provides reliable technical support for the accurate mapping and efficient application of 3D tiles and BIM vector model editing.

[0178] Based on these attributed component objects, a spatial index tree is constructed to accelerate spatial queries. A common implementation is to build an R-tree or a variant thereof. For example, the mapping and association system calculates the bounding box of each component object (such as a wall, panel, door, or window), which is the smallest cuboid that can completely enclose the component. Then, following the R-tree construction algorithm, bounding boxes of components with similar spatial locations are recursively grouped to form higher-level bounding boxes, ultimately forming a hierarchical tree-like index structure. The leaf nodes of this tree store references to component objects, while the intermediate nodes store the combined bounding boxes of the spaces covered by their child nodes. Traversing this spatial index tree, the system quickly filters out the set of components that need to be loaded based on the viewpoint position and field of view, thus generating tile load data. For example, when a user overlooks the entire building model, the system quickly locates the bounding box at the building outline level through the index tree and loads only simplified versions of the roof and exterior wall components; when the user zooms in to view the interior of a room, the mapping and association system delves into the leaf nodes of the index tree and loads all the detailed furniture and equipment components in that room. To achieve smooth switching between different precision levels, a coordinate transformation interface needs to be established. This interface is responsible for converting the vertex coordinates in the local coordinate system within a component object to coordinates in the global geographic coordinate system, based on its actual geographic location in Earth space. Then, according to the rules of the tile pyramid, it maps these coordinates to a specific level of tile mesh. Through this series of operations, the underlying tile binding from the original component data to customized 3D tiles that can be efficiently scheduled and rendered in the 3D engine is completed.

[0179] Furthermore, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment includes step S500:

[0180] Step S510: Locate the customized 3D tile according to the interaction command, and extract the unique node identifier from the underlying binding relationship of the tile.

[0181] The unique node identifier is derived using the following formula:

[0182] (25);

[0183] In formula (25), Indicates a unique node identifier. This indicates the identifier extraction function. This indicates the underlying binding relationship of the tile's underlying layer. The control logic of formula (25) is to locate the customized 3D tile based on the interaction command, extract the unique node identifier of the target node from its underlying binding relationship, realize the accurate mapping from the interaction operation to the unique node identifier, and provide a reliable identifier foundation for subsequent editing and querying; through the underlying binding relationship, a bidirectional association between 3D tiles and BIM components is established, supporting the reverse operation from tile to BIM; as the core interaction entry of the 3D tile and BIM vector model editing mapping association method, it provides reliable technical support for efficient and accurate BIM editing and scene interaction.

[0184] When a user selects a 3D tile of interest via interactive commands such as clicking a specific area on the screen, the system first locates the corresponding customized 3D tile based on the coordinate information in the command. These customized 3D tiles are pre-generated and contain local area data of the building model, such as the pier of a city bridge. Next, unique node identifiers are extracted from the underlying binding relationships of the tile's underlying structure. This binding relationship is essentially a data association table that links the tile to nodes in the underlying raster model. Specifically, the mapping association system queries the tile's metadata file, which records the node identifier corresponding to each tile. These identifiers are unique strings, such as "BridgePillar_001," used to identify a specific node of a bridge pillar. Through this extraction process, the mapping association system ensures an accurate correspondence from user interaction to data retrieval, avoiding confusion between different components in complex models.

[0185] Step S520: Obtain the raster voxel cluster and attribute feature vector containing voxel topology connection information through the unique node identifier.

[0186] The unique node identifier obtained is used to retrieve the relevant raster voxel clusters and attribute feature vectors from the database. A raster voxel cluster refers to a group of spatially connected voxel units. These voxel units are represented by topological connectivity information such as adjacency relationships. For example, in a bridge model, a cluster of voxels might represent the concrete portion of a pier. Its topological connectivity information describes the shared edges or faces between voxels, ensuring that voxels within the cluster form a continuous structure. The attribute feature vector is a multi-dimensional array containing business attributes such as "concrete strength grade C40," "pouring date 2023-05-15," and "anti-corrosion coating type." The mapping and association system queries the index database using the node identifier to quickly retrieve this information. For example, for "BridgePillar_001," it returns a voxel cluster dataset, where each voxel has location coordinates and connection labels, along with an attribute vector for subsequent shape processing.

[0187] Step S530: Analyze the raster voxel clusters and combine them with attribute feature vectors to determine the boundary constraints. Input the shape restoration algorithm to fit and obtain the vector vertex coordinate sequence.

[0188] The vector vertex coordinate sequence is obtained by fitting the data using the following formula:

[0189] (26);

[0190] In formula (26), This represents the sequence of vector vertex coordinates after fitting. This represents the sequence of vector vertex coordinates before fitting. This represents the rasterized output of the vertex sequence from the shape restoration algorithm. Represents raster data, Indicates the boundary constraint region. This represents minimizing the solution. The control logic of formula (26) is to fit an accurate sequence of vector vertex coordinates by minimizing the error between the vector rasterization result and the target raster voxel cluster under boundary constraints. By minimizing the error during fitting, a high-precision restoration from raster to vector is achieved, solving the geometric distortion problem of raster-vector conversion. Combined with the boundary constraints of attribute features, it ensures that the fitting result conforms to the engineering semantics of BIM components (such as size and type). As the core conversion step of 3D tile editing mapping to BIM vector model, it provides reliable algorithm support for "raster editing → vector restoration".

[0191] The process of resolving raster voxel clusters involves traversing all voxels within a cluster, analyzing their topological connectivity to identify boundaries, such as detecting whether the voxel's outer face is adjacent to other clusters or is empty space, thereby determining boundary constraints such as curvature limits or smoothness requirements. These conditions, combined with attribute feature vectors, such as the influence of strength level on boundary stiffness, are input into a shape reconstruction algorithm. This shape reconstruction algorithm is a least-squares fitting method that generates continuous curves from discrete points of voxels. For example, by fitting the cylindrical shape of a bridge pier, it outputs a vector vertex coordinate sequence, such as an ordered set of (x, y, z) points representing the pier's outline.

[0192] Step S540: Construct a geometric topological surface and map attribute feature vectors based on the fitted vector vertex coordinate sequence to generate the edited vector model.

[0193] The edited vector model is generated using the following formula:

[0194] (27);

[0195] In formula (27), This represents the edited vector model. Represents the original model. Represents the attribute feature vector. Represents a mapping function. Indicates the number of feature vectors. The tensor product is represented by the tensor product. The control logic of formula (27) is to generate the edited vector model by fusing new geometric topology and attribute features through tensor product on the basis of the original vector model. Through tensor product fusion, the geometric topology and attribute features are updated synchronously, avoiding data inconsistency. Based on the original model, only the edited part is superimposed, realizing efficient incremental update. As the final output of the 3D tile and BIM vector model editing mapping association method, it provides reliable technical support for the closed loop of "tile editing → BIM model update".

[0196] When constructing geometric topological surfaces using vector vertex coordinate sequences, the system connects these points to form a triangular mesh, for example, converting the vertex sequence of a bridge pier into a closed surface model. Then, attribute feature vectors are mapped onto this mesh, such as adding a "corrosion-resistant coating type" label to each facet, ultimately generating an edited vector model that supports further modifications such as resizing.

[0197] Furthermore, the mapping and association method for editing 3D tiles and BIM vector models proposed in this embodiment includes step S600 as follows:

[0198] Step S610: Obtain the edited vector model and calculate the geometric bounding box. Traverse the preset 3D tile spatial index tree to retrieve the tile encoding sequence to be updated that has spatial intersection with the geometric bounding box.

[0199] The tile encoding sequence to be updated is obtained using the following formula:

[0200] (28);

[0201] In formula (28), This represents the tile encoding sequence to be updated, which is an ordered set of all 3D tile encodings that require incremental updates. This refers to tile encoding, used to uniquely identify a single 3D tile, such as Z-order (Morton code). This represents a spatial index tree for three-dimensional tiles, used to organize and retrieve spatial index structures for three-dimensional tiles, such as quadtrees and octrees. This indicates a traversal operation to query the geometric bounding box. As a condition, traverse the spatial index tree. The operation. This indicates the result of the spatial intersection judgment, which determines whether two bounding boxes have a spatial intersection and returns true or false. This indicates the query's geometric bounding box, which is the bounding box of the edited vector model and serves as the range for the spatial query. The bounding box represents the spatial boundary of the 3D tile, used to determine whether it intersects with the query range. This formula is used to select the set of tile codes that intersect with the geometric bounding box. The control logic of formula (28) is to traverse the 3D tile spatial index tree, filter out the tile codes that have spatial intersection with the geometric bounding box of the edited model, form the sequence to be updated, and accurately lock the tiles affected by the editing through spatial index and intersection judgment, avoiding unnecessary full updates; the update range is reduced from the entire scene to local tiles, which greatly improves the update efficiency of 3D tiles after BIM editing; only the updated tile data is distributed, reducing network transmission and storage overhead.

[0202] Spatial intersection judgment result This can be derived from the following formula:

[0203] (29);

[0204] In formula (29), This indicates the result of the spatial intersection judgment. This indicates a query for the geometric bounding box. Indicates a box surrounded by tiles. This represents the coordinate axis dimension index. In a three-dimensional context, if the dimensions overlap, then there is an intersection. Represents three-dimensional logical AND, Indicates querying the geometric bounding box in Minimum value of the axis, Indicates the tile enclosure in The maximum value of the axis. Indicates the tile enclosure in Minimum value of the axis, Indicates querying the geometric bounding box in The maximum value of the axis. The control logic of formula (29) is to determine whether the two bounding boxes overlap for each coordinate axis dimension in three-dimensional space. Only when all dimensions overlap is it determined that there is a spatial intersection. By judging one by one through the three-dimensional dimensions, the accuracy of the spatial intersection judgment is ensured and misjudgment is avoided. Only simple comparison operation is required, and the computational complexity is low, which is suitable for fast query in large-scale tile scenarios. As the core judgment basis for the selection of tiles to be updated, it provides reliable algorithm support for the accurate incremental update of 3D tiles and BIM vector model editing.

[0205] Tile Enclosure Box This can be derived from the following formula:

[0206] (30);

[0207] In formula (30), The tile bounding box is defined by the axis-aligned bounding box (AABB) of each dimension's extreme values, which precisely represents the spatial extent of the tile. This represents the edited vector model point set, which is the set of coordinates of all vertices in the vector model. Represents the coordinates of a point in a vector model, or the coordinates of a point in three-dimensional space, such as... . Represents the minimum value of each dimension of the tile bounding box, point set exist The minimum values ​​on each axis constitute the minimum corner points of the bounding box. Represents the maximum value of each dimension of the tile bounding box, point set exist The maximum value on each axis constitutes the maximum corner point of the bounding box. The control logic of formula (30) is to define the minimum axis-aligned bounding box that can completely enclose the vector model by calculating the extreme values ​​of the vector model point set in each dimension. By calculating the extreme values, the minimum bounding box is defined, avoiding overestimation of the spatial range. It only requires traversing the point set and comparing the extreme values, and the computational complexity is O(N). It is suitable for large-scale vector model scenarios as the core basis for spatial intersection judgment and spatial indexing, and provides a reliable geometric basis for accurate incremental updates of 3D tiles and BIM vector models.

[0208] Based on these tile encoding sequences to be updated, the mapping and association system extracts the corresponding original tile loads from storage, which include the tile's geometric data and texture information. For example, for "Tile_001", the original tile load contains the original mesh data of the pier. The mapping and association system then calculates the geometric and topological differences between the edited vector model and the original loads, identifying changed parts by comparing vertex positions and face connections, such as a pier height increase of 10 units, thereby determining the local reconstruction region. This region is limited to the subspace with significant differences, avoiding global recalculation.

[0209] Step S620: Extract the corresponding original tile load based on the tile encoding sequence to be updated, and calculate the geometric topological difference between the edited vector model and the original tile load to determine the local reconstruction area.

[0210] The original tile load is obtained using the following formula:

[0211] (31);

[0212] In formula (31), This represents the original tile load, and the set of original geometry, attributes, and topology data corresponding to the tile to be updated, which serves as the benchmark for differential calculation. This represents the tile encoding sequence to be updated, which is a set of three-dimensional tile codes generated by formula (28) that requires incremental updates. This indicates an extraction operation, which retrieves and returns the corresponding original tile load from the tile storage or index based on the tile code. This formula is used to obtain the corresponding original tile load from the tile code sequence to be updated. The control logic of formula (31) is to accurately extract the original load data of the corresponding tile from the storage system based on the tile code sequence to be updated, and to directly index the original load data through the tile code, thereby achieving accurate extraction of the original load data and avoiding full scanning; only the load of the tile to be updated is extracted, which greatly reduces the amount of data reading and processing and improves the update efficiency; as a pre-step for difference calculation and local reconstruction, it provides a reliable data foundation for efficient incremental updates of 3D tiles and BIM vector model editing.

[0213] The local reconstruction region is determined using the following formula:

[0214] (32);

[0215] In formula (32), This represents the local reconstruction region, which is the collection of all tile region elements that need to be locally reconstructed. Represents a tile region element, a subdivided region unit within a tile, such as a sub-tile, voxel, or geometric facet. This represents the geometric and topological differences between the edited vector model and the original tile loads, and calculates the differences between the edited vector model and the original tile loads in the region. Geometric topological differences. The threshold value represents the difference threshold used to determine whether reconstruction is required. If the difference exceeds this value, reconstruction is deemed necessary. This formula is used to determine the local reconstruction areas that need to be reconstructed based on geometric topological differences. The control logic of formula (32) is based on geometric topological differences, filtering out tile areas with differences exceeding the preset threshold to form a set of local areas that need reconstruction. This narrows the update range from the entire tile to local areas, achieving a finer-grained incremental update. Through threshold filtering, unnecessary reconstruction of areas with no differences is avoided, significantly improving update performance. As the core step of tile incremental update, it provides reliable technical support for the efficient and accurate mapping of 3D tiles and BIM vector models.

[0216] Geometric and topological differences between the edited vector model and the original tile load This can be derived from the following formula:

[0217] (33);

[0218] In formula (33), This indicates the geometric and topological differences between the edited vector model and the original tile load, in the region. The average geometric difference between the edited vector model and the original tile load. Indicates the size of the point set, region The total number of 3D points (vertices) used for difference calculation. This indicates the edited vector model. The coordinates of each point, after editing, the vector model in the region Inner The three-dimensional coordinates of each vertex. Indicates the original tile load. The coordinates of the points, the original tile load in the region Inner The three-dimensional coordinates of each vertex. The Euclidean norm is used to calculate the Euclidean distance between two 3D points and quantify the geometric offset of a single vertex. The control logic of formula (33) is to quantify the geometric topological difference between the edited vector model and the original tile load by calculating the average Euclidean distance of the corresponding vertices in the region. The average Euclidean distance objectively and quantitatively expresses the degree of geometric topological change, avoiding subjective judgment. It only requires traversing the vertices and performing simple distance calculations, with a complexity of O(N3), which is suitable for large-scale tile scenarios. As the core basis for local reconstruction region selection, it provides reliable algorithm support for efficient and accurate incremental updates of 3D tiles and BIM vector models.

[0219] Once the mapping and association system receives the edited vector model, it first calculates its geometric bounding box, a minimum axis-aligned rectangle that encloses the model for fast spatial lookup. For example, in an editing scenario of a city bridge model, assuming the user modifies the shape of the piers, the system calculates the bounding box boundary values ​​based on the model's vertex coordinates, such as minimum coordinates (100, 200, 50) and maximum coordinates (150, 250, 100). These values ​​define the model's extent in 3D space. Next, the mapping and association system traverses a pre-defined 3D tile spatial index tree, a hierarchical tree structure, such as an octree, used to organize and manage tile data, where each node represents a spatial partition. By comparing the bounding box with the boundaries of nodes in the tree, the system retrieves tiles with spatial intersections, such as the tile encoding sequences identified as "Tile_001" and "Tile_002," which cover the pier area, ensuring that updates are only applied to the relevant parts.

[0220] Step S630: The vector geometric data in the local reconstruction area is converted into binary stream data blocks that conform to the current tile level resolution using the discretization voxel algorithm. The binary stream data blocks are used to replace the original data segments corresponding to the tile encoding sequence to be updated, while the data structure of the unchanged area is retained to generate the reconstructed 3D tile file.

[0221] The reconstructed 3D tile file is obtained using the following formula:

[0222] (34);

[0223] In formula (34), This represents the reconstructed 3D tile file, the final tile data file after the partial update is completed. This represents the original tile encoding sequence to be updated, which is the original 3D tile file data to be updated. Indicates the starting position of the original data segment to be replaced, which is the index of the starting byte in the original tile data that needs to be replaced. This indicates the end position of the original data segment to be replaced plus 1, and is the index of the end byte in the original tile data that needs to be replaced (following the left-closed, right-open principle). This represents a binary stream data block that conforms to the current tile level resolution, a new data block converted from the vector geometric data of the locally reconstructed region. The control logic of formula (34) is to precisely replace the specified data segment in the original tile file with the binary stream data block, while retaining the remaining data structure, to generate the incrementally updated tile file. By precisely replacing the data segment, the full tile rewriting is avoided, significantly improving the update efficiency; the data structure of the unchanged area is retained, ensuring the compatibility and stability of the tile file; as the final output of the 3D tile and BIM vector model editing mapping association method, it provides complete technical support for efficient and reliable incremental updates.

[0224] Binary stream data blocks that conform to the current tile level resolution This can be derived from the following formula:

[0225] (35);

[0226] In formula (35), This represents a binary stream data block that conforms to the current tile level resolution, used to replace the original tile data segment with encoded and compressed binary data. Indicates based on the current level For the current voxelization results The function performs encoding and compression operations on voxelized geometric data according to the tile level, achieving a match between detail and resolution. This represents the voxelization result, which is the voxel data transformed from vector geometric data within the local reconstruction region by the discretization voxel algorithm. The layer of the current tile is indicated. The layer of the tile in the spatial index tree determines the resolution and level of detail of the data. The control logic of formula (35) is to encode and compress the voxelized geometric data according to the current tile layer to generate binary stream data blocks that conform to the resolution of the layer. Through layer encoding, it is ensured that the resolution of the tile data is completely matched with the current view layer, avoiding data redundancy or loss of detail. Encoding and compression greatly reduce the size of the data block and improve storage and transmission efficiency. As the core data generation step for incremental tile updates, it provides reliable technical support for efficient and accurate mapping of 3D tiles and BIM vector model editing.

[0227] Current voxelization results It is obtained through the following formula:

[0228] (36);

[0229] In formula (36), This represents the current voxelization result, a binary voxel grid. This indicates that the current voxel is covered by local geometric data. Indicates an empty space. This represents a point in vector geometry data, a vertex in vector geometry, or a sampling point. This represents the set of vector geometric data (vertices, faces) within the local reconstruction region to be processed. . This represents the discretized voxel grid coordinates at the current tile level, and the index coordinates of the voxel in the 3D mesh. Indicates the coordinates projected onto the voxel grid. Within, it determines whether a geometric point or surface is located inside the target voxel. The control logic of formula (36) is to traverse the discretized voxel mesh, generate a binary voxelization result based on the projection coverage of the local vector geometric data, and achieve accurate mapping from vector geometry to voxel mesh through projection judgment, ensuring that the voxelization result accurately reflects the local geometric shape; the resolution of the voxel mesh is bound to the current tile level, ensuring a balance between data volume and level of detail; as a pre-step for generating binary stream data blocks, it provides a reliable geometric transformation basis for efficient incremental updates of 3D tiles and BIM vector models.

[0230] A discretized voxel algorithm is employed to process the local reconstruction region, converting vector geometric data into voxel form. Specifically, this discretized voxel algorithm first divides the region into a uniform grid, such as voxel units with a resolution of 1x1x1. Then, it samples the vector curves; for example, the cylindrical surface of a bridge pier is discretized into filled voxels, generating a binary stream data block, where 1 represents occupied and 0 represents idle. This binary stream data block conforms to the current tile layer resolution, such as the fineness of layer 10. This binary stream data block is used to replace the original data segment, while preserving the structure of unchanged areas such as the bridge deck, ultimately generating a reconstructed 3D tile file. This allows for efficient loading of local updates during rendering.

[0231] Step S640: Extract the attribute feature vectors of the edited vector model, establish a new key-value pair mapping between the attribute feature vectors and the voxel indexes in the reconstructed 3D tile file through the unique node identifier, write the new key-value pair mapping into the global attribute association table, complete the synchronous update of the attribute mapping relationship, and output the updated mapping association system containing the locally regenerated tiles.

[0232] The updated global attribute association table is derived using the following formula:

[0233] (37);

[0234] In formula (37), This represents the updated global attribute association table, the final global attribute association table after attribute mapping synchronization is completed, used to maintain the correspondence between tile voxels and BIM attributes. This represents the original global attribute association table, the global attribute association table before the edit operation, and contains historical key-value pair mappings. This represents a new key-value pair mapping, a set of new mapping relationships established by unique node identifiers, attribute feature vectors, and reconstructed tile voxel indices. The union operation merges the new key-value pair mappings into the original association table. If there is a key conflict, the new mapping overwrites the old mapping. The control logic of formula (37) is to merge the newly established attribute-voxel key-value pair mappings into the original global attribute association table, complete the synchronous update of the attribute mapping relationship, and ensure that the global attribute association table is completely synchronized with the latest tile and BIM model status through the union of sets, thus avoiding data inconsistency. The key-value pair structure ensures the efficiency of attribute query and supports fast attribute retrieval in large-scale scenarios. As the core data structure of the mapping association system, it provides reliable attribute management support for the closed loop of 3D tile and BIM vector model editing.

[0235] New key-value pair mapping This can be derived from the following formula:

[0236] (38);

[0237] In formula (38), This represents a new key-value pair mapping used to update the set of key-value pairs in the global attribute association table. This represents a unique node identifier, a globally unique node identifier, which is usually extracted from formula (25). Representing attribute feature vectors, from the edited vector model The extracted attribute information, such as material, color, ID, type, etc. Represents voxel indexes, and the reconstructed 3D tile file. The index of the corresponding voxel is used to locate geometric data. The control logic of formula (38) uses the unique node identifier as the key to bind the attribute feature vector and the corresponding voxel index as the value, and establishes a new key-value pair mapping. Through the unique node identifier, the precise binding of attributes, voxels and BIM nodes is realized, avoiding mapping ambiguity. The value contains both attributes and voxel index, taking into account semantic information and geometric positioning, and supports rich queries and operations. As the core data unit for updating the global attribute association table, it provides a reliable mapping basis for the closed loop of editing 3D tiles and BIM vector models.

[0238] The system extracts attribute feature vectors from the edited vector model, such as "concrete strength C50" and "coating thickness 2mm," and establishes a key-value mapping with voxel indices using the unique node identifier "BridgePillar_002." Specifically, the mapping association system associates vectors with voxels in the file, such as indices 100-200, and writes this information to a global attribute association table, a database table used to maintain the synchronization between attributes and geometry. After mapping is complete, the updated mapping association system is output, including locally regenerated tiles, thereby enabling real-time editing and attribute consistency of the bridge model, improving data accuracy and interaction efficiency in engineering simulations.

[0239] Please see Figure 2 This invention provides a mapping and association system for editing 3D tiles and BIM vector models, used to execute the aforementioned mapping and association method for editing 3D tiles and BIM vector models. It includes a grid model generation module 10, a node attribute set determination module 20, a node association relationship set acquisition module 30, a tile bottom layer binding completion module 40, a vector model acquisition module 50, and an attribute mapping relationship update module 60. The grid model generation module 10 is used to integrate the geometric information and topological relationships of the building information model vector data to generate a grid model; the node attribute set determination module 20 is used to parse the node structure of the grid model to extract attribute data, store the attribute data in a database, and determine the node attribute set; node association... The relationship set acquisition module 30 is used to generate unique identifiers for nodes in the node attribute set and establish a mapping relationship table, and obtain the node association relationship set based on the mapping relationship table; the tile bottom layer binding completion module 40 is used to extract basic component units from the raster model based on the node association relationship set, and generate customized 3D tiles by combining attribute mapping to complete the tile bottom layer binding; the vector model acquisition module 50 is used to obtain node identifiers through reverse lookup of tile bottom layer binding, and process the raster data using a restoration algorithm to obtain the edited vector model; the attribute mapping relationship update module 60 is used to locally regenerate 3D tiles and update the attribute mapping relationship based on the modified data of the edited vector model to obtain the updated mapping relationship system.

[0240] This embodiment provides a mapping and association method and system for editing 3D tiles and BIM vector models. Compared with the prior art, the beneficial effects achieved are as follows:

[0241] 1. The core logic of this embodiment is the forward conversion of "vector → raster → tile" + the reverse mapping of "tile editing → vector restoration", which establishes a precise association between 3D tiles and BIM vector models through unique identifiers and mapping relationship tables.

[0242] 2. The key innovations lie in the "tile bottom layer binding" and "local update strategy," which not only achieves two-way synchronization between visual editing and the native model, but also takes into account the editing efficiency of large-scale BIM models.

[0243] 3. The closed-loop process ensures the consistency of the BIM vector model after 3D tile editing, solving the core pain points of "asynchronous editing and incompatibility of attributes" in traditional BIM+3D tile applications.

[0244] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.

Claims

1. A mapping and association method for editing 3D tiles and BIM vector models, characterized in that, Includes the following steps: S100: Integrate the geometric information and topological relationships of building information model vector data to generate a raster model; S200. Analyze the node structure of the raster model to extract attribute data, store the attribute data in the database, and determine the node attribute set; S300: Generate unique identifiers for the nodes in the node attribute set and establish a mapping relationship table; obtain the node association relationship set according to the mapping relationship table. S400. Based on the set of node associations, extract basic component units from the grid model and combine them with attribute mapping to generate customized three-dimensional tiles to complete the bottom layer binding of the tiles. S500: Obtain the node identifier by reverse querying the underlying binding relationship of the tile bottom layer, and process the raster data using the restoration algorithm to obtain the edited vector model; S600. Based on the modified data of the edited vector model, locally regenerate the 3D tiles and update the attribute mapping relationship to obtain the updated mapping association system.

2. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, The building information model vector data includes the vector boundary data of the building information model components. Step S100 includes: S110. Obtain the vector boundary data of the building information model components, parse the vertex coordinates of the vector boundaries, and establish a set of geometric feature descriptions. S120. Based on the geometric feature description set, identify the topological adjacency and inclusion relationships between components and construct a global topological relationship graph; S130. Discretize the coverage area of ​​the global topological relationship map to generate an initial voxel mesh structure; S140. If the voxel grid in the initial voxel grid structure intersects with the vector boundary, then the attribute is mapped to the voxel grid. S150. Aggregate the mapped voxel mesh and output a raster model that integrates geometric information and topological relationships.

3. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, Step S200 includes: S210. Obtain a grid model that integrates geometric information and topological relationships, and use an octree depth-first traversal algorithm to scan the grid model to extract the original attribute data sequence. S220. Parse the original attribute data sequence and convert it into structured attribute records according to the field mapping rules; S230. Write the structured attribute records into a database storage table to generate a persistent attribute data table. S240. Traverse the field metadata of the persistent attribute data table, and aggregate the non-empty field names and data type definitions in the field metadata to determine the node attribute set.

4. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, Step S300 includes: S310. Traverse the node attribute set to extract spatial location coordinates and attribute feature data, and use a hash algorithm to combine the spatial location coordinates and attribute feature data to generate a globally unique identifier. S320. Construct a node index mapping table based on the globally unique identifier, wherein the node index mapping table contains the correspondence between node index keys and original storage addresses; S330. Retrieve the node index mapping table to obtain the neighbor node index keys of the current node, and parse the spatial topology connection rules between the neighbor nodes and the current node; S340. If the spatial distance between the spatial location vector of a neighboring node and the spatial location vector of the current node is less than or equal to a preset distance threshold, and the connection relationship between the spatial location vector of the neighboring node and the spatial location vector of the current node satisfies a preset connectivity condition, then a node adjacency tuple is generated, and all the node adjacency tuples are aggregated to obtain a set of node association relationships.

5. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, Step S400 includes: S410. Generate a raster voxel cluster by aggregating voxel units based on the node association relationship set, wherein the node association relationship set represents the voxel topological connection; S420. Construct a component entity mesh for the grid voxel cluster, and map the attribute feature vector to the component entity mesh to generate a component object with attributes; S430. Construct a spatial index tree based on the component object with attributes, traverse the spatial index tree to generate tile load data and establish a coordinate transformation interface to complete the bottom layer binding of the customized three-dimensional tile.

6. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, Step S500 includes: S510: Locate customized 3D tiles based on interactive commands, and extract unique node identifiers from the underlying binding relationships of the tile's bottom layer; S520. Obtain the raster voxel cluster and attribute feature vector containing voxel topology connection information through the unique node identifier; S530. Analyze the raster voxel cluster and combine it with the attribute feature vector to determine the boundary constraint conditions, and input the shape restoration algorithm to fit the vector vertex coordinate sequence; S540. Construct a geometric topological surface based on the fitted vector vertex coordinate sequence and map the attribute feature vector to generate the edited vector model.

7. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 1, characterized in that, Step S600 includes: S610. Obtain the edited vector model and calculate the geometric bounding box. Traverse the preset three-dimensional tile space index tree to retrieve the tile encoding sequence to be updated that has a spatial intersection with the geometric bounding box. S620. Extract the corresponding original tile load according to the tile encoding sequence to be updated, and calculate the geometric topological difference between the edited vector model and the original tile load to determine the local reconstruction region; S630. The vector geometric data in the local reconstruction area is converted into binary stream data blocks that conform to the current tile level resolution using the discretization voxel algorithm. The binary stream data blocks are used to replace the original data segments corresponding to the tile encoding sequence to be updated, while retaining the data structure of the unchanged area to generate the reconstructed three-dimensional tile file. S640. Extract the attribute feature vectors of the edited vector model, establish a new key-value pair mapping between the attribute feature vectors and the voxel indexes in the reconstructed 3D tile file through a unique node identifier, write the new key-value pair mapping into the global attribute association table, complete the synchronous update of the attribute mapping relationship, and output the updated mapping association system containing locally regenerated tiles.

8. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 7, characterized in that, In step S610, the tile encoding sequence to be updated is obtained using the following formula: ； in, This indicates the tile encoding sequence to be updated. Indicates tile encoding, Represents a three-dimensional tile spatial index tree. This indicates a traversal operation. This indicates the result of the spatial intersection judgment. This indicates a query for the geometric bounding box. This refers to a box surrounded by tiles.

9. The mapping and association method for editing 3D tiles and BIM vector models as described in claim 8, characterized in that, In step S610, the spatial intersection judgment result This can be derived from the following formula: ； in, This indicates the result of the spatial intersection judgment. This indicates a query for the geometric bounding box. Indicates a box surrounded by tiles. This represents the coordinate axis dimension index. In a three-dimensional context, if the dimensions overlap, then there is an intersection. Represents three-dimensional logical AND, Indicates querying the geometric bounding box in Minimum value of the axis, Indicates the tile enclosure in The maximum value of the axis. Indicates the tile enclosure in Minimum value of the axis, Indicates querying the geometric bounding box in The maximum value of the axis; Tile Enclosure Box This can be derived from the following formula: ； in, Indicates a box surrounded by tiles. This represents the edited vector model point set. Represents the coordinates of a point in a vector model. This represents the minimum value of each dimension of the tile bounding box. This represents the maximum value of each dimension of the tile bounding box.

10. A mapping and association system for editing 3D tiles and BIM vector models, used to execute the mapping and association method for editing 3D tiles and BIM vector models as described in any one of claims 1 to 9, characterized in that, include: The raster model generation module (10) is used to integrate the geometric information and topological relationships of the building information model vector data to generate a raster model; The node attribute set determination module (20) is used to parse the node structure of the raster model to extract attribute data, store the attribute data in the database, and determine the node attribute set; The node association set acquisition module (30) is used to generate unique identifiers for nodes in the node attribute set and establish a mapping relationship table, and to acquire the node association set according to the mapping relationship table; The tile bottom layer binding completion module (40) is used to extract basic component units from the grid model based on the node association relationship set, and generate customized three-dimensional tiles by combining attribute mapping to complete the tile bottom layer binding; The vector model acquisition module (50) is used to obtain node identifiers through reverse querying of the tile bottom layer binding and to process raster data using a restoration algorithm to obtain the edited vector model. The attribute mapping relationship update module (60) is used to locally regenerate three-dimensional tiles and update attribute mapping relationships based on the modified data of the edited vector model, so as to obtain the updated mapping relationship system.

Images