A sectioning method and system for complex BIM models of rail transit stations

Abstract

The present invention proposes a cutting method and system for complex BIM models of rail transit stations, and relates to the technical field of cutting BIM models of rail transit stations. The method includes: reconstructing multi-format initial BIM model data into a hierarchical structure data set that retains geometric topology and attribute integrity; establishing a spatial index architecture based on the data set, generating ordered index intervals by mapping coordinates through Z-order curves, calculating the three-dimensional influence domain range in combination with the cutting plane position and retrieving the intersecting spatial element set; intersecting the element set with the cutting plane to generate a geometric intersection point, and forming a closed polygonal ring contour through adjacency analysis to aggregate it into a cutting surface; truncating the model based on the cutting surface, hiding the specified side geometry and reconstructing the topology, and generating a visual cutting result through material mapping rendering. It solves the problems of multi-source data compatibility conflicts, low global traversal efficiency and insufficient continuity of special-shaped structures, and improves the BIM cutting accuracy and efficiency of rail transit stations.

Description

Technical Field

[0001] The present invention relates to the technical field of sectioning of a rail transit station BIM model, and in particular to a sectioning method and system for a complex BIM model of a rail transit station. Background Art

[0002] With the rapid development of urban rail transit construction, the application of BIM technology in rail transit station projects is becoming increasingly in-depth. However, when faced with the complex BIM models of rail transit stations that integrate ultra-large scale and multiple disciplines, existing sectioning technologies have the following key flaws: First, multi-format compatibility conflicts. Disciplines such as architecture, structure, and electromechanical use BIM models in different formats, such as .rvt and .dgn. During integration, these conflicts can lead to misaligned sections and missing components due to coordinate system deviations, inconsistent design standards across disciplines, and insufficient data compatibility between cross-disciplinary models. Second, global traversal efficiency is low. Traditional sectioning requires traversing tens of thousands of components (such as beams, columns, and pipelines), resulting in memory overload and computational delays, making it impossible to meet the needs of real-time interaction. Third, the lack of continuity in special-shaped structures: When sectioning curved structures such as station domes and special-shaped entrances and exits, insufficient mesh discretization results in jagged faults or incomplete contours, compromising geometric integrity.

[0003] Therefore, there is an urgent need for a rail transit station BIM model cutting method that is compatible with multi-source data, avoids global traversal, and ensures the cutting accuracy of special-shaped structures. Summary of the Invention

[0004] In view of the above problems existing in the prior art, the first aspect of the present invention proposes a sectioning method for a complex BIM model of a rail transit station, comprising:

[0005] Step 1: Reconstruct the initial BIM model data in different formats into a unified hierarchical structure dataset, preserving the geometric topological features, attribute data, and relationship data integrity of the initial BIM model data;

[0006] Step 2: Establish a spatial index structure based on the hierarchical structure dataset, calculate the 3D influence domain based on the spatial index structure and the position of the cutting plane, and retrieve the set of spatial voxels that intersect with the cutting plane within the 3D influence domain;

[0007] Step 3: Intersect the spatial element set with the cutting plane to generate a set of geometric intersection points. Based on the set of geometric intersection points, a closed polygonal ring contour is generated through adjacency analysis and aggregated into a cutting surface.

[0008] Step 4: Truncate the initial BIM model based on the cutting plane, hide the model geometry on the specified side of the cutting plane, reconstruct the truncated BIM model geometry and render it to generate a visual cutting result.

[0009] In conjunction with the first aspect, in some implementations, reconstructing the hierarchical structured dataset in step 1 includes:

[0010] Step 1-1: Based on the initial BIM model, generate a pre-processed model by removing redundant geometric elements and unifying the spatial coordinate system;

[0011] Step 1-2: Based on the engineering logic of the rail transit station, a hierarchical framework is generated by establishing a five-level spatial hierarchy of the station as a whole, floor areas, functional areas, component groups, and single components;

[0012] Steps 1-3: Based on the preprocessing model and hierarchical framework, generate a hierarchical mapping relationship by parsing the component space attributes and mapping them to corresponding nodes;

[0013] Steps 1-4: Based on the hierarchical mapping relationship, the spatial position, geometric data and attributes of the components are stored through nodes to generate a hierarchical structure data set.

[0014] In combination with the first aspect, in some implementations, the floor area includes the station hall floor and the platform floor; the functional area includes at least the public area, the equipment area and the office area; the component group includes at least the air-conditioning unit, the smoke exhaust system, the access control system and the power distribution system; the single component includes at least the heater, the air circuit breaker, the fan and the distribution box.

[0015] In conjunction with the first aspect, in some implementations, establishing the spatial index architecture in step 2 includes:

[0016] Step 2-1: Based on the 3D space of the initial BIM model, a set of discretized spatial voxels is generated by constructing a cube grid and performing dynamic subdivision processing;

[0017] Step 2-2: Based on the discretized spatial voxel set, generate a spatial voxel dataset with coordinates by mapping the physical entity and recording the center coordinates;

[0018] Step 2-3: Based on the spatial voxel dataset with coordinates, map the coordinates to linear index codes using a space-filling curve to generate an index code set;

[0019] Step 2-4: Generate an ordered spatial voxel index interval based on the index coding set through sorting operations.

[0020] In conjunction with the first aspect, in some implementations, retrieving the spatial voxel set in step 2 includes:

[0021] Step 2-5: Generate the 3D influence domain boundary by calculating the minimum intersecting bounding box based on the position of the cutting plane;

[0022] Step 2-6: Generate index code sub-intervals based on the vertices of the three-dimensional influence domain boundary through space filling curve encoding;

[0023] Step 2-7: Generate a candidate voxel set through local range query based on the ordered spatial voxel index interval and index code subinterval;

[0024] Step 2-8: Based on the candidate voxel set and the cutting plane, the spatial intersection relationship is calculated by the directed distance sign judgment method to generate a set of spatial voxels that intersect with the cutting plane.

[0025] In conjunction with the first aspect, in some implementations, the intersection calculation in step 3 includes:

[0026] Step 3-1: Generate a vertex distance symbol set based on the vertex coordinates of the spatial voxel and the cutting plane equation through directed distance calculation;

[0027] Step 3-2: Based on the vertex distance sign set, generate a set of edges to be intersected by traversing the edges of the voxels and detecting edges with opposite signs;

[0028] Step 3-3: Based on the set of edges to be intersected, calculate the coordinates of the intersection points by linear interpolation to generate a set of geometric intersection points.

[0029] In conjunction with the first aspect, in some implementations, generating a closed polygonal annular contour in step 3 and aggregating it into a cutting surface includes:

[0030] Step 3-4: Based on the geometric intersection set, generate a single contour intersection sequence by traversing unvisited intersections and tracing adjacent points;

[0031] Step 3-5: Based on the single contour intersection sequence, generate a closed circular contour or an unclosed contour identifier through end-to-end closure verification;

[0032] Step 3-6: Based on the unclosed contour identification, generate a forced closed contour through interpolation completion operation;

[0033] Step 3-7: Perform polygon aggregation operation on the closed circular contour and the forced closed contour to generate a cutting surface.

[0034] In conjunction with the first aspect, in some implementations, hiding the model geometry on the specified side of the cutting plane in step 4 includes:

[0035] Step 4-1: Based on the initial BIM models on both sides of the cutting plane, generate a set of direction vectors by selecting spatial elements and calculating the vectors from the center points of the spatial elements to the cutting plane;

[0036] Step 4-2: Based on the direction vector set and the cutting plane normal vector, generate a dot product value symbol set through dot product operation;

[0037] Step 4-3: Based on the dot product value symbol set, generate the to-be-hidden side identifier by hiding the negative-value side spatial voxels.

[0038] In conjunction with the first aspect, in some implementations, reconstructing the truncated BIM model geometry and rendering it in step 4 to generate a visual sectioning result includes:

[0039] Step 4-4: Generate a new topological structure by updating the connection relationship between vertices, edges, and faces based on the visible model data after truncation and the cutting surface geometry data;

[0040] Step 4-5: Generate a revised topology model by checking and repairing the orientation consistency of the BIM model surface and the closedness of the BIM model body based on the new topology structure and the attribute data in the hierarchical structure dataset;

[0041] Step 4-6: Based on the modified topology model, generate a geometric model with material attributes by mapping the material to the cutting surface;

[0042] Steps 4-7: Based on the geometric model with material attributes, generate visual cutting results through vertex interpolation and rendering.

[0043] In a second aspect, the present invention provides a sectioning system for a complex BIM model of a rail transit station. The system adopts the method provided in any of the above embodiments, and the system includes:

[0044] Data reconstruction module, used to reconstruct the initial BIM model data in different formats into a unified hierarchical structure data set, preserving the geometric topological features, attribute data and relationship data integrity of the initial BIM model data;

[0045] The spatial index module is connected to the data reconstruction module and is used to establish a spatial index structure based on the hierarchical structure data set, calculate the three-dimensional influence domain range based on the spatial index structure and the position of the cutting plane, and retrieve the set of spatial elements that intersect the cutting plane within the three-dimensional influence domain;

[0046] The cutting surface generation module is connected to the spatial index module and is used to perform intersection calculation on the spatial element set and the cutting plane to generate a set of geometric intersection points. Based on the set of geometric intersection points, a closed polygonal ring contour is generated through adjacency analysis and aggregated into a cutting surface.

[0047] The model truncation module is connected to the cutting plane generation module and is used to truncate the initial BIM model based on the cutting plane, hide the model geometry on the specified side of the cutting plane, reconstruct the geometric shape of the truncation BIM model and render it to generate a visual cutting result.

[0048] Compared with the prior art, the beneficial effects of the present invention are: Step 1 resolves compatibility conflicts by reconstructing the dataset, reconstructing the initial BIM model data in different formats into a unified hierarchical structure dataset, retaining the geometric topological features, attribute data and relationship data integrity, eliminating coordinate system deviation and data loss during multi-professional model integration, and avoiding section dislocation from the source; Step 2 improves efficiency by establishing an index architecture, establishes a spatial index architecture based on the hierarchical structure dataset, calculates the three-dimensional influence domain range in combination with the sectioning plane position, optimizes the full model traversal into a local spatial voxel set retrieval, reduces invalid calculations, and solves memory overload and delay problems; Step 3 ensures continuity by generating a closed-loop contour, intersects the spatial voxel set with the sectioning plane to generate a geometric intersection set, forms a closed polygonal ring contour based on the adjacency relationship analysis and aggregates it into a cutting surface, ensuring that there is no fault at the cutting edge of the special-shaped structure; Step 4 realizes visualization through integrated truncation rendering, hides the specified side geometry after truncating the initial BIM model based on the cutting surface, reconstructs the topology and renders to generate a visual sectioning result, forming a complete technical closed loop.

[0049] In summary, the unified dataset in step 1 provides a structured basis for indexing in step 2; the efficient retrieval in step 2 compresses the intersection calculation range in step 3 to the influence domain; the closed contour generated in step 3 directly supports the precise truncation rendering in step 4, ultimately achieving efficient, accurate, and continuous BIM sectioning of rail transit stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 The figure shows a flow chart of a sectioning method for a complex BIM model of a rail transit station provided by an embodiment of the present invention.

[0052] Figure 2 Shown is a schematic diagram of a BIM model of a rail transit station provided by an embodiment of the present invention.

[0053] Figure 3 Shown is a schematic diagram of a partial fragment of a divided spatial hierarchical structure provided by an embodiment of the present invention.

[0054] Figure 4 Schematic diagram of an initial spatial voxel and a spatial voxel after X / Y / Z three-axis synchronous binary division operation provided by an embodiment of the present invention.

[0055] Figure 5 Shown is a schematic diagram of a cut-away BIM model provided by an embodiment of the present invention.

[0056] Figure 6 The figure shows a structural schematic diagram of a sectioning system for a complex BIM model of a rail transit station provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0058] The specific embodiments of the present invention are described below.

[0059] Example 1

[0060] Combine Figures 1 to 5 As shown, the present invention proposes a sectioning method for a complex BIM model of a rail transit station, comprising:

[0061] Step 1: Reconstruct the initial BIM model data in different formats into a unified hierarchical structure dataset, preserving the geometric topological features, attribute data, and relationship data integrity of the initial BIM model data.

[0062] Specifically, the initial BIM model may contain data in various formats, including .rvt, .rfa, .dgn, .sat, .stp, .gltf, and .obj. Using a lossless data conversion mechanism, redundant geometric elements (such as isolated points, lines, and surfaces) are first removed and geometric errors (such as inconsistent normal directions or missing meshes) are corrected. The local coordinate systems of each discipline model are aligned to the global coordinate system, eliminating section misalignment caused by coordinate system deviations. Subsequently, a five-level spatial hierarchy is constructed based on the engineering logic of the rail transit station: the entire station, floor areas (such as the concourse and platform levels), functional areas (such as public areas, equipment areas, and office areas), component groups (such as air conditioning units and smoke exhaust systems), and individual components (such as fans and distribution boxes). The spatial attributes of the components in the initial BIM model (such as spatial coordinates and geometric dimensions) are analyzed and mapped to corresponding hierarchical nodes, ultimately generating a hierarchical structured dataset that retains complete geometric topology and attribute associations. This step eliminates compatibility conflicts when integrating multiple discipline models, ensuring data consistency in subsequent sectioning operations.

[0063] Step 2: Establish a spatial index structure based on the hierarchical structure dataset, calculate the three-dimensional influence domain range based on the spatial index structure and the position of the cutting plane, and retrieve the set of spatial elements that intersect with the cutting plane within the three-dimensional influence domain.

[0064] Specifically, in order to avoid memory overload caused by global traversal of tens of thousands of components, a spatial index architecture is established based on a hierarchical structured data set. Specifically: a cube grid of initial size is constructed in three-dimensional space, and a discretized spatial voxel set is generated through dynamic subdivision processing. If the voxel is completely outside the model, it is discarded; if it is completely inside, the subdivision is terminated; if it intersects with the model, the X / Y / Z three-axis synchronous binary division is performed until the size is less than the threshold. After recording the center coordinates of each voxel, the three-dimensional Z-order curve is used to reduce the dimension of the coordinates and map them into a linear index code (for example, the coordinate (4,5,6) is binary interleaved as 111100010, corresponding to the decimal index code 482), and an ordered spatial voxel index interval [K min , K max When cutting, the minimum intersection bounding box is calculated according to the position of the cutting plane to determine the three-dimensional influence domain range [x min , x max ]×[y min ,y max ]×[z min , z max ], convert its boundary vertices into index code subintervals [k min , k max A local range query is used to obtain a set of candidate voxels. Then, a directed distance sign judgment method is used to accurately select the set of spatial voxels that actually intersect the cutting plane. This step compresses the calculation scope to the influence area near the cutting plane, reducing invalid traversals and addressing memory overload and real-time interaction delays.

[0065] Step 3: Perform intersection calculation on the spatial element set and the cutting plane to generate a geometric intersection set. Based on the geometric intersection set, generate a closed polygonal ring contour through adjacency analysis and aggregate it into a cutting surface.

[0066] Specifically, the intersection of the set of spatial voxels and the cutting plane is calculated: the signed distances of the voxel vertices are calculated based on the plane equation, the edges are traversed and the edges with opposite signs are detected, the intersection coordinates are calculated using the linear interpolation formula, and a set of geometric intersections is generated. Subsequently, a closed contour is generated through adjacency analysis: starting from unvisited intersections, adjacent points are traced to form a single contour intersection sequence. If the beginning and end are not closed (such as the open boundary of the voxel edge), the interpolation is completed to a forced closed contour. Multiple closed circular contours are aggregated to form a complete cutting surface. This process ensures the accuracy of the surface structure by dynamically subdividing the voxels, and avoids the jagged faults of special-shaped structures such as domes in traditional methods through a forced closure mechanism, ensuring geometric continuity.

[0067] Step 4: Truncate the initial BIM model based on the cutting plane, hide the model geometry on the specified side of the cutting plane, reconstruct the truncated BIM model geometry and render it to generate a visual cutting result.

[0068] Specifically, the initial BIM model is truncated based on the cutting plane: the dot product of the direction vector from the voxel center to the cutting plane and the plane normal is calculated, and the model geometry on the side with a negative dot product value is hidden. The vertex, edge, and face connectivity of the truncated model is updated, and surface orientation consistency and model closure are checked and repaired. The cutting plane inherits the original component's material properties (such as color and texture), and a smooth transition is achieved through vertex interpolation. Finally, the visual sectioning result is rendered. This step completes the technical loop from data reconstruction to result output.

[0069] The embodiments of the present invention eliminate multi-format compatibility conflicts and avoid sectioning dislocations by unifying hierarchical data sets; the spatial index architecture optimizes global traversal into local influence domain retrieval, improving computational efficiency; dynamic subdivision of voxels and forced closed contours ensure the accuracy and continuity of sectioning of special-shaped structures; material inheritance and topology repair ensure the engineering practicality of the visualization results, forming an efficient, accurate, and complete technical closed loop.

[0070] In conjunction with the first aspect, in some implementations, reconstructing the hierarchical structured dataset in step 1 includes:

[0071] Step 1-1: Based on the initial BIM model, generate a pre-processed model by removing redundant geometric elements and unifying the spatial coordinate system.

[0072] Specifically, the initial BIM model must remove redundant geometric elements (such as duplicate vertices and unreferenced edges) and correct geometric errors (such as non-manifold edges or self-intersecting faces). Unifying the spatial coordinate system is crucial: converting the local coordinate systems of specialized models, such as architectural, structural, and electromechanical, to a global coordinate system. For example, aligning the relative coordinates of the .rvt file with the absolute coordinates of the .dgn file eliminates component misalignment caused by coordinate system deviations. This step provides a geometrically pure and spatially consistent preprocessing model for establishing the hierarchical framework.

[0073] Step 1-2: Based on the engineering logic of the rail transit station, a hierarchical framework is generated by establishing a five-level spatial hierarchy of the station as a whole, floor area, functional area, component group, and single component.

[0074] Specifically, the system is divided into several levels based on the engineering logic of rail transit stations: the overall station, encompassing the building boundaries; floor areas, including physical layers such as the concourse and platform levels; functional areas, such as public areas (gates, platforms), equipment areas (machine rooms), and office areas (control rooms); component groups, divided by system, such as air conditioning units (including fans and coils) and access control systems (including card readers and controllers); and single components, the smallest units such as distribution boxes and air circuit breakers. This framework replaces the traditional loose geometric organization approach by associating components through engineering semantics.

[0075] Step 1-3: Based on the preprocessing model and hierarchical framework, generate a hierarchical mapping relationship by parsing the component space attributes and mapping them to the corresponding nodes.

[0076] Specifically, the spatial properties of components in the preprocessing model (such as coordinate range and geometry type) are analyzed and mapped to corresponding hierarchical nodes. For example, a fan located in the equipment area of ​​the station hall is mapped to the "Station Hall - Equipment Area - Air Conditioning Unit - Fan" node to ensure that the components strictly correspond to the engineering logic.

[0077] Steps 1-4: Based on the hierarchical mapping relationship, the spatial position, geometric data and attributes of the components are stored through nodes to generate a hierarchical structure data set.

[0078] Specifically, each hierarchical node stores three types of data: spatial location (such as bounding box coordinates), geometric data (such as mesh vertices), and attributes (such as material and model). Nodes are linked through parent-child relationships, forming a hierarchical dataset that combines geometric precision with engineering semantics.

[0079] Among them, functional areas can be expanded to commercial areas, evacuation passages, etc.; component groups can include professional groups such as water supply and drainage systems, fire protection systems, etc.

[0080] In the embodiment of the present invention, redundancy removal reduces storage load, and the coordinate system eliminates integration deviation; the five-level hierarchical mapping engineering logic improves data manageability; attribute inheritance ensures lossless transmission of BIM information; and provides a structured basis for spatial indexing.

[0081] In combination with the first aspect, in some implementations, the floor area includes the station hall floor and the platform floor; the functional area includes at least the public area, the equipment area and the office area; the component group includes at least the air-conditioning unit, the smoke exhaust system, the access control system and the power distribution system; the single component includes at least the heater, the air circuit breaker, the fan and the distribution box.

[0082] Specifically, the floor areas are: the concourse level (including ticketing and security functions) and the platform level (including waiting and boarding areas), which are connected by stairs and escalator components;

[0083] Functional areas: public areas (such as station distribution halls), equipment areas (such as environmental control rooms), and office areas (such as dispatching rooms). The equipment areas are usually adjacent to the platform level to reduce the distance for laying pipelines.

[0084] Component group: air-conditioning unit (including fan, heater), smoke exhaust system (including air damper, smoke detector), access control system (including card reader, electric lock), power distribution system (including distribution cabinet, circuit breaker);

[0085] Single components: such as coils in air conditioning systems and air circuit breakers in power distribution systems.

[0086] In an embodiment of the present invention, the hierarchical division matches the station operation and maintenance logic, improving the engineering readability of the sectioning results; the component group classification supports system-level sectioning analysis (such as separately sectioning the smoke exhaust duct); and avoids component mismapping caused by unstructured data.

[0087] In conjunction with the first aspect, in some implementations, establishing the spatial index architecture in step 2 includes:

[0088] Step 2-1: Based on the three-dimensional space of the initial BIM model, a set of discretized spatial elements is generated by constructing a cube grid and performing dynamic subdivision processing.

[0089] Specifically, a cube mesh of initial size is constructed in the 3D space of the BIM model. The initial size of the voxels is set according to the model accuracy requirements (such as surface curvature), and dynamic subdivision is performed:

[0090] If the voxel is completely inside the model, terminate the subdivision and keep it;

[0091] If it is completely outside, discard it;

[0092] If it intersects the model, it is bisected synchronously along the X / Y / Z axes to generate eight sub-voxels, and this process is repeated until the voxel size is smaller than a threshold (e.g., the minimum surface approximation error is met). This step balances accuracy and computational efficiency through adaptive subdivision.

[0093] Step 2-2: Based on the discretized spatial voxel set, generate a spatial voxel dataset with coordinates by mapping the physical entity and recording the center coordinates.

[0094] Specifically, the discretized physical entities (such as beams, columns, and pipelines) are mapped to intersecting elements, and each element stores the entity's geometric data and records the center point coordinates.

[0095] Step 2-3: Based on the spatial voxel dataset with coordinates, map the coordinates to linear index codes through a space filling curve to generate an index code set.

[0096] Specifically, the voxel center coordinates are converted to binary and interleaved. For example, the coordinates (4,5,6) (binary 100,101,110) are interleaved to 111100010, which converts to the decimal index 482. This operation preserves spatial proximity: adjacent voxels in 3D space remain adjacent in the 1D index.

[0097] Step 2-4: Generate an ordered spatial voxel index interval based on the index coding set through sorting operations.

[0098] Specifically, sort all index codes and generate a continuous interval [K min , K max ], supporting efficient range queries.

[0099] Among them, the space filling curve can use the Hilbert curve instead of the Z-order curve, sacrificing some efficiency in exchange for higher spatial aggregation.

[0100] In the embodiment of the present invention, dynamic discretization optimizes storage to avoid memory waste of uniform grids; Z-order mapping reduces the dimensionality of multi-dimensional queries to one dimension to accelerate retrieval; and spatial proximity retention ensures the effectiveness of local traversal.

[0101] In conjunction with the first aspect, in some implementations, retrieving the spatial voxel set in step 2 includes:

[0102] Step 2-5: Generate the 3D influence domain boundary by calculating the minimum intersecting bounding box based on the position of the cutting plane.

[0103] Specifically, the minimum intersecting bounding box within the model range is calculated according to the cutting plane equation (such as Ax+By+Cz+D=0), and the axially aligned bounding box (AABB) boundary [x min , x max ]×[y min ,y max ]×[z min , z max ].

[0104] Step 2-6: Generate index code sub-intervals based on the vertices of the three-dimensional influence domain boundary through space filling curve encoding.

[0105] Specifically, for the minimum vertex of the bounding box (x min ,y min , z min ) and the maximum vertex (x max ,y max , z max ) are Z-order encoded to generate index code sub-intervals [k min , k max ].

[0106] Step 2-7: Based on the ordered spatial voxel index interval and the index code sub-interval, generate a candidate voxel set through local range query.

[0107] Specifically, in the global index interval [K min , K max ] the query belongs to the subinterval [k min , k max ] to generate a set of candidate voxels and avoid full model scanning.

[0108] Step 2-8: Based on the candidate voxel set and the cutting plane, the spatial intersection relationship is calculated by the directed distance sign judgment method to generate a set of spatial voxels that intersect with the cutting plane.

[0109] Specifically, because the Z-order curve may cover non-intersecting areas (such as the corners of the bounding box), the signed distance judgment method is used to filter the candidate voxels twice: the spatial relationship between the voxel bounding box and the cutting plane is calculated, and only the voxels that actually intersect are retained.

[0110] In the embodiment of the present invention, the minimum bounding box compresses the search range; the fuzzy search utilizes the orderliness of the index to improve efficiency; and the secondary filtering ensures the accuracy of the result and avoids false elimination.

[0111] In conjunction with the first aspect, in some implementations, the intersection calculation in step 3 includes:

[0112] Step 3-1: Generate a vertex distance symbol set based on the vertex coordinates of the spatial voxel and the cutting plane equation through directed distance calculation;

[0113] Step 3-2: Based on the vertex distance sign set, generate a set of edges to be intersected by traversing the edges of the voxels and detecting edges with opposite signs;

[0114] Step 3-3: Based on the set of edges to be intersected, calculate the coordinates of the intersection points by linear interpolation to generate a set of geometric intersection points.

[0115] Specifically, the signed distance value of each vertex of the spatial voxel cube is calculated based on the cutting plane equation (form Ax + By + Cz + D = 0). The sign of this distance value represents the vertex's orientation relative to the cutting plane: a positive value indicates the vertex is located in front of the cutting plane, a negative value indicates it is located behind it, and a value of zero indicates the vertex is directly on the plane. This symbolic judgment provides the basis for subsequent edge screening.

[0116] Next, we traverse the 12 edges of the spatial element and check whether the distances between the vertices at each end of the edge have opposite signs. If so, the edge must intersect the cutting plane; otherwise, it does not. This step efficiently selects all edges to be intersected, avoiding unnecessary calculations.

[0117] Finally, perform linear interpolation on each edge to be intersected. Assume that the edge endpoints P1 (x1, y1, z1) and P2 (x2, y2, z2) have opposite distances. The intersection parameter t is calculated using the formula: , calculate the actual intersection coordinates based on the parameter t: , after traversing all the edges to be intersected, a complete set of geometric intersection points is generated.

[0118] In an embodiment of the present invention, the signed distance sign method quickly locates intersecting edges and reduces computational complexity; linear interpolation ensures the accuracy of intersection coordinates and avoids jagged errors caused by grid discretization; and a complete set of geometric intersections provides a data basis for closed contour generation.

[0119] In conjunction with the first aspect, in some implementations, generating a closed polygonal annular contour in step 3 and aggregating it into a cutting surface includes:

[0120] Step 3-4: Based on the geometric intersection set, generate a single contour intersection sequence by traversing unvisited intersections and tracing adjacent points;

[0121] Step 3-5: Based on the single contour intersection sequence, generate a closed circular contour or an unclosed contour identifier through end-to-end closure verification;

[0122] Step 3-6: Based on the unclosed contour identification, generate a forced closed contour through interpolation completion operation;

[0123] Step 3-7: Perform polygon aggregation operation on the closed circular contour and the forced closed contour to generate a cutting surface.

[0124] Specifically, first, an unvisited intersection point is selected from the geometric intersection set as the starting point. This point is then traced based on adjacency relationships to form a single-contour intersection sequence. Specifically, within the intersection plane of the starting voxel, adjacent intersection points are searched clockwise and added to the sequence until no unvisited adjacent points are found or the process returns to the starting point. This step establishes topological associations between intersection points based on voxel adjacency.

[0125] Next, the intersection sequence is verified for closure. If the sequence's beginning and end coincide and contain at least three intersection points, a closed circular contour is generated. If it's not closed (commonly seen at model boundaries or surface discrete gaps), a forced closure mechanism is activated: interpolation points are inserted between the beginning and end points to complete the path and ensure the contour is closed. This operation completely solves the problem of contour fragmentation caused by mesh discretization in traditional sectioning.

[0126] Subsequently, the above process is repeated until all intersection points are visited, generating multiple independent closed contours (e.g., a dome surface may produce multiple annular contours).

[0127] Finally, polygon aggregation is performed on all closed loop contours: adjacent contours are merged into a single continuous cutting surface through triangulation or Boolean operations.

[0128] In the embodiment of the present invention, adjacent point tracking maintains the correctness of contour topology; a forced closing mechanism eliminates sectioning of irregular structures; and polygon aggregation forms a complete cutting surface to support accurate model truncation.

[0129] In conjunction with the first aspect, in some implementations, hiding the model geometry on the specified side of the cutting plane in step 4 includes:

[0130] Step 4-1: Based on the initial BIM models on both sides of the cutting plane, generate a set of direction vectors by selecting spatial elements and calculating the vectors from the center points of the spatial elements to the cutting plane;

[0131] Step 4-2: Based on the direction vector set and the cutting plane normal vector, generate a dot product value symbol set through dot product operation;

[0132] Step 4-3: Based on the dot product value symbol set, generate the to-be-hidden side identifier by hiding the negative-value side spatial voxels.

[0133] First, select the complete voxels of the initial BIM model on both sides of the cutting surface (the integrity must not be destroyed by cutting), and calculate the direction vector from its center point to the cutting plane. Assume that the coordinates of the voxel center are O(x0, y0, z0), then the direction vector Defined by the formula: , where (x p ,y p , z p ) is any reference point on the cutting plane.

[0134] Next, get the normal vector of the cutting plane = (A, B, C) (determined by the plane equation coefficients), calculate the dot product of the direction vector and the normal vector: .

[0135] Finally, a hiding operation is performed based on the sign of the dot product value:

[0136] like <0, hide all model geometry on the side of the element;

[0137] like >0, hide the opposite side model geometry;

[0138] like =0 (the center of the voxel is exactly on the plane) is processed according to the preset strategy (such as hiding one side by default).

[0139] In the embodiment of the present invention, the dot product operation is mathematically clear and computationally efficient; complete voxel selection avoids misjudgment near the cutting surface; and symbol judgment accurately separates the models on both sides of the section.

[0140] In conjunction with the first aspect, in some implementations, reconstructing the truncated BIM model geometry and rendering it in step 4 to generate a visual sectioning result includes:

[0141] Step 4-4: Generate a new topological structure by updating the connection relationship between vertices, edges, and faces based on the visible model data after truncation and the cutting surface geometry data;

[0142] Step 4-5: Generate a revised topology model by checking and repairing the orientation consistency of the BIM model surface and the closedness of the BIM model body based on the new topology structure and the attribute data in the hierarchical structure dataset;

[0143] Step 4-6: Based on the modified topology model, generate a geometric model with material attributes by mapping the material to the cutting surface;

[0144] Steps 4-7: Based on the geometric model with material attributes, generate visual cutting results through vertex interpolation and rendering.

[0145] First, the model's topology is updated: Based on the visible model data after truncation and the geometry of the cut surface, the vertex-edge-face connections are reconstructed. The newly added cut surface vertices are re-meshed with the original vertices, ensuring that each edge connects to the correct vertices and each face is enclosed by a closed boundary. This step resolves topological breaks caused by the cut.

[0146] Secondly, check and repair model topology defects: adjust the normal direction of the triangular facets so that they are uniformly outward, complete open boundaries caused by sectioning (such as equipment pipeline cuts), and ensure that the repair conforms to engineering logic by introducing attribute data from the hierarchical structure data set (such as component connection relationships).

[0147] Subsequently, material properties are assigned to the cutting surface, inheriting the color and texture properties of the original component (such as gray for concrete structures and silver for stainless steel equipment), and a smooth material transition between the cutting surface and adjacent surfaces is achieved through vertex attribute interpolation.

[0148] Finally, rendering optimization is performed, using a lighting model to enhance the three-dimensional sense of volume, and LOD (level of detail) control based on the spatial hierarchy structure to improve the rendering efficiency of large-scale scenes. The generated visual sectioning results retain both geometric accuracy and engineering semantic information.

[0149] In the embodiment of the present invention, topology reconstruction ensures the validity of geometric data; attribute inheritance maintains the continuity of BIM information; material interpolation eliminates visual mutations; and hierarchical rendering optimization improves the smoothness of interaction.

[0150] Example 2

[0151] like Figure 6 As shown, in a second aspect, the present invention provides a sectioning system for a complex BIM model of a rail transit station. The system adopts the method provided in any of the above embodiments, and the system includes:

[0152] Data reconstruction module, used to reconstruct the initial BIM model data in different formats into a unified hierarchical structure data set, preserving the geometric topological features, attribute data and relationship data integrity of the initial BIM model data;

[0153] The spatial index module is connected to the data reconstruction module and is used to establish a spatial index structure based on the hierarchical structure data set, calculate the three-dimensional influence domain range based on the spatial index structure and the position of the cutting plane, and retrieve the set of spatial elements that intersect the cutting plane within the three-dimensional influence domain;

[0154] The cutting surface generation module is connected to the spatial index module and is used to perform intersection calculation on the spatial element set and the cutting plane to generate a set of geometric intersection points. Based on the set of geometric intersection points, a closed polygonal ring contour is generated through adjacency analysis and aggregated into a cutting surface.

[0155] The model truncation module is connected to the cutting plane generation module and is used to truncate the initial BIM model based on the cutting plane, hide the model geometry on the specified side of the cutting plane, reconstruct the geometric shape of the truncation BIM model and render it to generate a visual cutting result.

[0156] The system corresponds to the method provided in Example 1 and will not be described in detail here.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A sectioning method for a complex BIM model of a rail transit station, characterized in that: include: Step 1: Reconstruct the original BIM model data in different formats, which integrates the architecture, structure, and electromechanical multi-disciplinary models of the rail transit station, into a unified hierarchical structure dataset, preserving the geometric topological features, attribute data, and relationship data integrity of the original BIM model data; Step 2: Establishing a spatial index structure based on the hierarchical structure data set, calculating the range of a three-dimensional influence domain according to the spatial index structure and the position of the cutting plane, and retrieving a set of spatial voxels intersecting the cutting plane within the three-dimensional influence domain; Step 3: performing intersection calculation on the set of spatial elements intersecting with the cutting plane and the cutting plane to generate a set of geometric intersection points, and generating a closed polygonal annular contour based on the set of geometric intersection points through adjacency analysis, and aggregating them into a cutting surface; Step 4: truncating the initial BIM model based on the cutting plane, hiding the model geometry on the specified side of the cutting plane, reconstructing the truncated BIM model geometry and rendering it to generate a visual cutting result; The step 1 of reconstructing the hierarchical structure data set includes: Step 1-1: Based on the initial BIM model, generate a pre-processed model by removing redundant geometric elements and unifying the spatial coordinate system; Step 1-2: Based on the engineering logic of the rail transit station, a hierarchical framework is generated by establishing a five-level spatial hierarchy of the station as a whole, floor areas, functional areas, component groups, and single components; Steps 1-3: Based on the preprocessing model and hierarchical framework, generate a hierarchical mapping relationship by parsing the component space attributes and mapping them to corresponding nodes; Steps 1-4: Based on the hierarchical mapping relationship, the spatial position, geometric data and attributes of the components are stored through nodes to generate a hierarchical structure data set; The step 2 of establishing the spatial index architecture includes: Step 2-1: Based on the 3D space of the initial BIM model, a set of discretized spatial voxels is generated by constructing a cube grid and performing dynamic subdivision processing; Step 2-2: Based on the discretized spatial voxel set, generate a spatial voxel dataset with coordinates by mapping the physical entity and recording the center coordinates; Step 2-3: Based on the spatial voxel dataset with coordinates, map the coordinates to linear index codes using a space-filling curve to generate an index code set; Step 2-4: Generate an ordered spatial voxel index interval based on the index code set through sorting operation; Retrieving the spatial voxel set in step 2 includes: Step 2-5: Generate the 3D influence domain boundary by calculating the minimum intersecting bounding box based on the position of the cutting plane; Step 2-6: Generate index code sub-intervals based on the vertices of the three-dimensional influence domain boundary through space filling curve encoding; Step 2-7: Generate a candidate voxel set through local range query based on the ordered spatial voxel index interval and index code subinterval; Step 2-8: Based on the candidate voxel set and the cutting plane, the spatial intersection relationship is calculated by the directed distance sign judgment method to generate a set of spatial voxels that intersect with the cutting plane.

2. The sectioning method for a complex BIM model of a rail transit station according to claim 1, characterized in that: The floor area includes the station hall floor and the platform floor; the functional area includes at least the public area, equipment area and office area; the component group includes at least the air-conditioning unit, smoke exhaust system, access control system and power distribution system; the single component includes at least the heater, air circuit breaker, fan and distribution box.

3. The sectioning method for a complex BIM model of a rail transit station according to claim 1, characterized in that: The intersection calculation in step 3 includes: Step 3-1: Generate a vertex distance symbol set based on the vertex coordinates of the spatial voxel and the cutting plane equation through directed distance calculation; Step 3-2: Based on the vertex distance sign set, generate a set of edges to be intersected by traversing the edges of the voxels and detecting edges with opposite signs; Step 3-3: Based on the set of edges to be intersected, calculate the coordinates of the intersection points by linear interpolation to generate a set of geometric intersection points.

4. The sectioning method for a complex BIM model of a rail transit station according to claim 1, characterized in that: The step 3 generates a closed polygonal annular contour, which is aggregated into a cutting surface, including: Step 3-4: Based on the geometric intersection set, generate a single contour intersection sequence by traversing unvisited intersections and tracing adjacent points; Step 3-5: Based on the single contour intersection sequence, generate a closed circular contour or an unclosed contour identifier through end-to-end closure verification; Step 3-6: Based on the unclosed contour identification, generate a forced closed contour through interpolation completion operation; Step 3-7: Perform polygon aggregation operation on the closed circular contour and the forced closed contour to generate a cutting surface.

5. The sectioning method for a complex BIM model of a rail transit station according to claim 1, characterized in that: The model geometry on the specified side of the cutting plane hidden in step 4 includes: Step 4-1: Based on the initial BIM models on both sides of the cutting plane, generate a set of direction vectors by selecting spatial elements and calculating the vectors from the center points of the spatial elements to the cutting plane; Step 4-2: Based on the direction vector set and the cutting plane normal vector, generate a dot product value symbol set through dot product operation; Step 4-3: Based on the dot product value symbol set, generate the to-be-hidden side identifier by hiding the negative-value side spatial voxels.

6. The sectioning method for a complex BIM model of a rail transit station according to claim 1, characterized in that: In step 4, the geometric shape of the truncated BIM model is reconstructed and rendered to generate a visual sectioning result, including: Step 4-4: Generate a new topological structure by updating the connection relationship between vertices, edges, and faces based on the visible model data after truncation and the cutting surface geometry data; Step 4-5: Generate a revised topology model by checking and repairing the orientation consistency of the BIM model surface and the closedness of the BIM model body based on the new topology structure and the attribute data in the hierarchical structure dataset; Step 4-6: Based on the modified topology model, generate a geometric model with material attributes by mapping the material to the cutting surface; Steps 4-7: Based on the geometric model with material attributes, generate visual cutting results through vertex interpolation and rendering.

7. A sectioning system for complex BIM models of rail transit stations, characterized by: The system adopts the method according to any one of claims 1 to 6, and the system includes: The data reconstruction module is used to reconstruct the initial BIM model data of different formats, which integrates the architectural, structural, and electromechanical multi-disciplinary models of the rail transit station, into a unified hierarchical structure data set, preserving the geometric topological characteristics, attribute data, and relationship data integrity of the initial BIM model data; a spatial index module, connected to the data reconstruction module, configured to establish a spatial index framework based on the hierarchical structure data set, calculate a three-dimensional influence domain range based on the spatial index framework and the position of the cutting plane, and retrieve a set of spatial voxels intersecting the cutting plane within the three-dimensional influence domain; a cutting plane generation module, connected to the spatial index module, configured to perform intersection calculation on the set of spatial voxels intersecting the cutting plane and the cutting plane to generate a set of geometric intersection points, and generate a closed polygonal annular contour based on the set of geometric intersection points through adjacency analysis to aggregate the contours into a cutting plane; a model truncation module, connected to the cutting plane generation module, configured to truncate the initial BIM model based on the cutting plane, hide the model geometry on the specified side of the cutting plane, reconstruct the geometric shape of the truncation BIM model, render it, and generate a visual cutting result; The reconstructed hierarchical structure data set includes: Step 1-1: Based on the initial BIM model, generate a pre-processed model by removing redundant geometric elements and unifying the spatial coordinate system; Step 1-2: Based on the engineering logic of the rail transit station, a hierarchical framework is generated by establishing a five-level spatial hierarchy of the station as a whole, floor areas, functional areas, component groups, and single components; Steps 1-3: Based on the preprocessing model and hierarchical framework, generate a hierarchical mapping relationship by parsing the component space attributes and mapping them to corresponding nodes; Steps 1-4: Based on the hierarchical mapping relationship, the spatial position, geometric data and attributes of the components are stored through nodes to generate a hierarchical structure data set; The spatial indexing architecture includes: Step 2-1: Based on the 3D space of the initial BIM model, a set of discretized spatial voxels is generated by constructing a cube grid and performing dynamic subdivision processing; Step 2-2: Based on the discretized spatial voxel set, generate a spatial voxel dataset with coordinates by mapping the physical entity and recording the center coordinates; Step 2-3: Based on the spatial voxel dataset with coordinates, map the coordinates to linear index codes using a space-filling curve to generate an index code set; Step 2-4: Generate an ordered spatial voxel index interval based on the index code set through sorting operation; The search space voxel set includes: Step 2-5: Generate the 3D influence domain boundary by calculating the minimum intersecting bounding box based on the position of the cutting plane; Step 2-6: Generate index code sub-intervals based on the vertices of the three-dimensional influence domain boundary through space filling curve encoding; Step 2-7: Generate a candidate voxel set through local range query based on the ordered spatial voxel index interval and index code subinterval; Step 2-8: Based on the candidate voxel set and the cutting plane, the spatial intersection relationship is calculated by the directed distance sign judgment method to generate a set of spatial voxels that intersect with the cutting plane.

Images