Three-dimensional map construction method and system for substation construction site

Abstract

This application provides a method and system for constructing 3D maps at substation construction sites, relating to the field of 3D map technology. The method includes: acquiring point cloud data of the construction site using LiDAR; obtaining corresponding image data using an image acquisition device; performing filtering, downsampling, and coordinate registration processing on the point cloud data; simultaneously performing distortion correction and texture extraction on the image data; aligning the point cloud data and image data by coordinates and identifying fusion features to generate a 3D model; integrating various structural data based on the 3D model to construct structured 3D update data; adding the structured 3D update data to the 3D model; verifying the parameters of the 3D map data; and fusing the verified 3D map parameters with the 3D model to generate a 3D map. This application can solve the technical problem of poor accuracy in 3D map construction in existing technologies, achieving the technical effect of improving the accuracy of 3D map construction.

Description

Technical Field

[0001] This application relates to the field of 3D map technology, and in particular to a method and system for constructing 3D maps for substation construction sites. Background Technology

[0002] With the continuous expansion of power infrastructure construction and the increasing demands for refined management and construction safety in substation projects, the spatial environment of construction sites is becoming increasingly complex. Traditional methods relying on manual surveying, two-dimensional drawings, and static model management are no longer sufficient to meet the real-time, accuracy, and three-dimensional visualization requirements of construction organization. Currently, existing technologies still suffer from insufficient texture detail and incomplete geometric structures in point cloud and image fusion, especially in areas with dense steel structures, where texture drift and registration error accumulation are more prominent. Furthermore, some technologies rely on manual classification and labeling of components, failing to automatically establish component-level structural data relationships, making it difficult to support subsequent construction monitoring data updates, attribute association mapping, and construction status tracking.

[0003] In summary, existing technologies suffer from the lack of a unified 3D map construction method capable of multi-phase, multi-view point cloud acquisition, texture fusion, component-level semantic segmentation, structured data updates, and dynamic construction navigation in substation construction scenarios. This results in an inability to accurately reflect the real-time structural status of the construction site, hindering data-driven construction decision support and further impacting the safety, execution efficiency, and level of refined management during the construction process. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for constructing 3D maps for substation construction sites, in order to solve the technical problems in the prior art where there is a lack of a unified 3D map construction method that can realize multi-phase and multi-view point cloud acquisition, texture fusion, component-level semantic segmentation, structured data update and dynamic construction navigation in substation construction scenarios. This results in the inability to accurately reflect the real-time structural status of the construction site, the inability to realize data-driven construction decision support, and further affects the safety, execution efficiency and refined management level of the construction process.

[0005] In view of the above problems, this application provides a method and system for constructing three-dimensional maps for substation construction sites.

[0006] Firstly, this application provides a method for constructing a 3D map for a substation construction site, implemented through a 3D map construction system for a substation construction site. The method includes: collecting point cloud data of the construction site using LiDAR; acquiring corresponding image data using an image acquisition device; filtering, downsampling, and coordinate registration of the point cloud data; simultaneously performing distortion correction and texture extraction on the image data; aligning the point cloud data with the image data using coordinates; identifying and fusing features to generate a 3D model with texture details; integrating various structural data based on the 3D model; receiving real-time construction monitoring data for status updates to construct structured 3D update data; adding the structured 3D update data to the 3D model; verifying the parameters of the 3D map data; and fusing the verified 3D map parameters with the 3D model to generate a 3D map.

[0007] Preferably, the method for constructing a 3D map for a substation construction site further includes: decomposing construction nodes based on substation construction to determine construction cycle nodes; locating construction objects in the construction area and construction targets according to the construction cycle nodes, and analyzing the acquisition perspectives based on the located construction objects to determine multiple acquisition perspectives to ensure no blind spots in coverage, and setting high-precision targets as the common alignment reference for multi-cycle, multi-angle point cloud registration; configuring multi-period, multi-angle laser acquisition plans based on the construction cycle nodes and the corresponding multiple acquisition perspectives, setting corresponding high-precision targets, and acquiring point cloud data of the construction site.

[0008] Preferably, the method for constructing a 3D map for a substation construction site further includes: aligning point cloud data with image data according to the acquisition target and acquisition coordinates; generating a multi-level triangular mesh model from macro to micro based on the original high-density point cloud; matching the texture source of visual angles using the normal vectors of the point cloud data and the image data based on the coordinate alignment relationship, and establishing a texture alignment relationship between the point cloud data and the image data; adding the texture alignment relationship to the multi-level triangular mesh model to establish texture coordinate alignment between the image data texture and the multi-level triangular network.

[0009] Preferably, the method for constructing a 3D map for a substation construction site further includes: performing multi-level feature fusion based on the texture coordinate alignment relationship between the image data texture and the multi-level triangular network, wherein the texture features of the image data are fused with the hierarchical structure features of the point cloud data; and constructing the 3D model with texture details according to the spatial coordinate relationship of the multi-level feature fusion.

[0010] Preferably, the method for constructing a three-dimensional map for a substation construction site further includes: performing semantic segmentation of texture features and point cloud structural features based on the three-dimensional model to determine different component objects; establishing structural data for each segmented component object as an independent query entity in the database; and adding the structural data of each component to the three-dimensional model.

[0011] Preferably, the method for constructing a three-dimensional map for a substation construction site further includes: based on the component objects in the three-dimensional model, performing construction-related structural analysis, establishing a component construction structure model, and creating a unique digital identity code and associated attribute set for each component object; based on the component objects and the component construction structure model, receiving real-time construction monitoring data, inserting data according to the association mapping relationship of the component objects, updating the current construction spatial structural state of the component objects and the component association attribute data, and constructing structured three-dimensional update data for the three-dimensional model.

[0012] Preferably, the method for constructing a three-dimensional map for a substation construction site further includes: selecting at least N precise coordinate control points and installation constraints for component objects from the three-dimensional model, wherein N is not less than 10; verifying the coordinate accuracy of the three-dimensional map data using the N precise coordinate control points to obtain coordinate verification results; verifying the construction space and assembly parameters of the three-dimensional map data using the installation constraints to obtain assembly verification results; and determining the three-dimensional map parameters and generating the three-dimensional map when both the coordinate verification results and the assembly verification results pass.

[0013] Preferably, the method for constructing a three-dimensional map for a substation construction site further includes: connecting to construction logs to obtain construction progress plans, material information, quality inspection reports, construction monitoring parameters, and material traceability codes; performing structured data transformation on the obtained construction log data to generate construction record data, storing it in the corresponding component object data, and updating the structured three-dimensional data of the three-dimensional model.

[0014] Preferably, the method for constructing a three-dimensional map for a substation construction site further includes: acquiring a construction schedule plan; analyzing the construction schedule plan for construction parameters, construction target components, and construction objectives; simulating the construction parameters and construction target components in the three-dimensional map; searching for the optimal construction strategy to achieve the construction objectives; and generating a construction navigation path; integrating the construction navigation path into the three-dimensional map for interactive construction visualization, used for real-time construction navigation guidance.

[0015] Secondly, this application also provides a 3D map construction system for substation construction sites, used to execute the 3D map construction method for substation construction sites as described in the first aspect, including: a data processing module, used to collect point cloud data of the construction site through lidar, acquire corresponding image data by image acquisition equipment, perform filtering, downsampling and coordinate registration processing on the point cloud data, and simultaneously perform distortion correction and texture extraction on the image data; a fusion feature recognition module, used to align the point cloud data and the image data with coordinates, and recognize fusion features to generate a 3D model with texture details; a 3D update data construction module, used to integrate various structural data based on the 3D model, and receive real-time construction monitoring data for status updates to construct structured 3D update data; and a 3D map generation module, used to add the structured 3D update data to the 3D model, verify the parameters of the 3D map data, and fuse the verified 3D map parameters with the 3D model to generate a 3D map.

[0016] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of digital modeling and dynamic updating of the entire life cycle, all components and all elements of the substation construction site, it achieves the technical effects of high-precision spatial restoration, real-time status perception, intelligent construction navigation guidance and full-process visual collaborative management in complex construction environments.

[0017] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the three-dimensional map construction method for substation construction sites used in this application.

[0020] Figure 2 This is a schematic diagram of the structure of the 3D map building system used in this application for substation construction sites.

[0021] Figure labeling: Data processing module 1, fusion feature recognition module 2, 3D update data construction module 3, 3D map generation module 4. Detailed Implementation

[0022] This application provides a method and system for constructing 3D maps for substation construction sites. It addresses the technical problem in existing technologies where the lack of a unified 3D map construction method capable of multi-phase, multi-view point cloud acquisition, texture fusion, component-level semantic segmentation, structured data updates, and dynamic construction navigation in substation construction scenarios leads to an inaccurate reflection of the real-time structural status of the construction site. This hinders data-driven construction decision support and further impacts the safety, efficiency, and refined management of the construction process. The application achieves the technical goal of full-cycle, full-component, and full-element digital modeling and dynamic updating of substation construction sites, enabling high-precision spatial reconstruction, real-time status perception, intelligent construction navigation guidance, and full-process visualized collaborative management in complex construction environments.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0024] Example 1, please refer to the appendix. Figure 1 This application provides a method for constructing a 3D map of a substation construction site, which is applied to a 3D map construction system for a substation construction site, and specifically includes the following steps: S1: Point cloud data of the construction site is collected by LiDAR, and corresponding image data is obtained by image acquisition equipment. The point cloud data is filtered, downsampled and registered, and the image data is distorted and texture extracted.

[0025] Furthermore, this application also includes: decomposing construction nodes based on substation construction to determine construction cycle nodes; locating construction objects in the construction area and construction targets according to the construction cycle nodes, and analyzing the acquisition perspectives based on the located construction objects to determine multiple acquisition perspectives to ensure coverage without blind spots, and setting high-precision targets as the alignment common reference for multi-cycle, multi-angle point cloud registration; configuring multi-period, multi-angle laser acquisition plans based on the construction cycle nodes and the corresponding multiple acquisition perspectives, setting corresponding high-precision targets, and acquiring point cloud data of the construction site.

[0026] Specifically, decomposing construction nodes based on substation construction and determining construction cycle nodes refers to dividing the overall substation construction process into several progress nodes with independent work content according to construction logic and sequence, when conducting overall time planning. Each node corresponds to a relatively stable construction state. Construction nodes are usually set based on engineering activities in stages such as civil engineering, installation, and commissioning, so that the construction scenarios of different stages can be recorded in segments during point cloud acquisition. For example, if a framework needs to be installed in a certain construction stage, this stage can be used as a separate cycle node to collect spatial structural changes in that stage, thereby providing a stable time reference for subsequent alignment, fusion, and modeling, and thus forming a periodic acquisition framework that runs through the entire construction process.

[0027] The location of construction objects based on construction cycle nodes refers to identifying the specific spatial area requiring construction operations and the components, equipment, or structures within that area that need to be collected during the construction time period corresponding to a determined construction node. Construction object location is based on existing design drawings, BIM information, or on-site construction markings. By analyzing the location of the objects, further analysis of the acquisition perspective can be performed. Acquisition perspective analysis involves calculating the appropriate LiDAR deployment location and shooting direction based on the dimensions of the construction object in three-dimensional space, occlusion conditions, and the surrounding environment to ensure that the surface of each object can be completely scanned during the actual acquisition process. Based on this, multiple acquisition perspectives are determined to achieve comprehensive coverage of the construction area. Furthermore, high-precision targets are set up to establish a common alignment benchmark between different perspectives and cycles. High-precision targets typically use stable reflective materials and precisely known three-dimensional coordinates to unify point cloud data collected at different times into the same coordinate system through registration between point clouds from multiple cycles and angles.

[0028] Based on construction cycle nodes and corresponding multi-view acquisition procedures, configuring a multi-phase, multi-view laser acquisition plan involves systematically planning the acquisition time, acquisition path, equipment location, and target deployment scheme of the lidar after clarifying the acquisition cycle and acquisition viewpoints, to form a complete acquisition plan. The acquisition plan includes multiple acquisition tasks to record structural changes at the construction site over multiple construction cycles; it also includes multi-view acquisition tasks to ensure sufficient coverage and redundancy in the scan data of each cycle. During the execution of the acquisition plan, corresponding high-precision targets need to be reset for each phase and each viewpoint location to ensure the feasibility of registration between point cloud data. Finally, point cloud data acquisition of the construction site is completed according to the acquisition plan, obtaining three-dimensional point cloud data that reflects the true spatial structure of the construction area.

[0029] Furthermore, acquiring corresponding image information through image acquisition equipment refers to using cameras, panoramic imagers, or other imaging devices deployed at the construction site to simultaneously or quasi-synchronously acquire images of the same spatial area of ​​the point cloud collected by LiDAR, in order to obtain raw image data reflecting the actual surface texture, color distribution, and detailed features of the site. Image acquisition equipment includes fixed cameras, cameras mounted on UAVs, or imaging modules on mobile measurement platforms. The acquired image information not only serves as the basic data for subsequent texture bonding but also assists in identifying the boundaries of target components in the point cloud data, improving the visual precision of the 3D model.

[0030] Filtering, downsampling, and coordinate registration of point cloud data refers to removing isolated and noisy points caused by abnormal reflections, occlusion changes, or environmental interference after acquiring the original point cloud, thus ensuring the validity of the point cloud. After filtering, downsampling is performed to reduce the amount of point cloud data, improve processing efficiency, and maintain the integrity of the model's main structure. Methods include voxel mesh downsampling, where different voxel densities can be set according to the structural complexity of the construction site, multiplying the number of sampling points to improve local representation. Coordinate registration unifies point cloud data from different viewpoints, periods, or devices into the same coordinate system. This is generally achieved through target registration, feature point matching, or iterative nearest-point algorithms, maintaining spatial consistency among multiple point clouds.

[0031] Simultaneous distortion correction and texture extraction of image data refers to geometrically correcting the image to address potential lens distortion in image acquisition devices. This eliminates edge stretching, bending, or proportional distortion, ensuring the image accurately reflects the actual spatial structure. The distortion-corrected image then uses texture extraction algorithms to obtain surface color, brightness gradients, and detailed texture information. This information is used to provide a realistic appearance for the triangular mesh during subsequent 3D modeling, ensuring both structural geometric accuracy and visual realism in the final model.

[0032] S2: Align the point cloud data with the image data in coordinates and identify fusion features to generate a 3D model with texture details.

[0033] Furthermore, this application also includes: aligning point cloud data with image data according to the acquisition target and acquisition coordinates; generating a multi-level triangular mesh model from macro to micro based on the original high-density point cloud; matching the texture source of visual angles using the normal vectors of the point cloud data and the image data based on the coordinate alignment relationship, and establishing a texture alignment relationship between the point cloud data and the image data; adding the texture alignment relationship to the multi-level triangular mesh model to establish texture coordinate alignment between the image data texture and the multi-level triangular network.

[0034] Furthermore, this application also includes: performing multi-level feature fusion based on the texture coordinate alignment relationship between the image data texture and the multi-level triangular network, wherein the texture features of the image data are fused with the hierarchical structure features of the point cloud data; and constructing the three-dimensional model with texture details according to the spatial coordinate relationship of the multi-level feature fusion.

[0035] Furthermore, this application also includes: performing semantic segmentation of texture features and point cloud structural features based on the three-dimensional model to determine different component objects; establishing structural data of each component as an independent query entity in the database for each segmented component object; and adding the structural data of each component to the three-dimensional model.

[0036] Specifically, coordinate alignment between point cloud data and image data, based on the acquisition target and coordinates, refers to unifying the coordinate differences between different data sources, considering the spatial coordinate system used by the LiDAR to acquire point clouds, the extrinsic and intrinsic parameters used by the camera to acquire images, and their spatial positions relative to the construction object during the actual acquisition process. Coordinate alignment includes spatial translation, rotation, and scaling adjustments to ensure that pixels in the image and spatial points in the point cloud can find an accurate correspondence within the same coordinate system.

[0037] Generating a multi-level triangular mesh model from macro to micro based on the original high-density point cloud refers to using high-precision point cloud as initial geometric data and constructing a multi-level 3D patch structure through a hierarchical meshing algorithm. The macro level is used to describe the general geometric outline of the entire construction scene, while the micro level is used to present local details requiring high precision, such as component edges and connection points. The multi-level triangular mesh model is formed through progressive refinement or simplification, so that the number of meshes in each level increases exponentially compared to the previous level, to adapt to the modeling needs of scenes with different precision requirements.

[0038] Based on coordinate alignment, texture source matching is performed using the normal vectors of point cloud data and image data to establish a texture alignment relationship between point cloud data and image data. This means that after the coordinates of the point cloud and image are unified, the most suitable image is selected as the texture source for that surface region based on the visual angle between the normal vector direction of the points on the point cloud surface and the camera's shooting direction. The smaller the visual angle, the more perpendicular the image can be to the target surface, and the clearer the texture details; therefore, such image regions are preferentially selected as texture data. When multiple images have candidate relationships with the same surface, the optimal texture source is determined through strategies such as visual angle comparison and texture clarity judgment, thus forming an accurate texture correspondence between the point cloud and the image.

[0039] Adding texture alignment relationships to multi-level triangular mesh models establishes texture coordinate alignment between image data textures and the multi-level triangular network. This involves mapping texture information to each vertex, edge, and face of the multi-level triangular mesh after texture source matching, and generating a corresponding texture coordinate system for each. This ensures that textures are displayed on the corresponding surfaces with the correct coordinate positions during 3D rendering. Texture coordinate alignment uses a two-dimensional texture coordinate mapping method to establish a one-to-one correspondence between pixel coordinates in the image and geometric coordinates in the triangular mesh, enabling the model to maintain visual consistency across different levels and levels of precision.

[0040] Multi-level feature fusion, based on the alignment of image data texture with the texture coordinates of a multi-level triangular network, refers to the process of integrating features from different data sources according to the hierarchical structure of the triangular mesh after achieving precise coordinate matching between the image texture and the triangular mesh model. Texture features refer to the visual information in image data that characterizes the brightness, color variations, texture coarseness, and edge morphology of a component surface; while hierarchical structure features refer to the geometric structure information contained in different precision levels formed during the point cloud mesh construction process, including macroscopic contours, mesoscopic structural divisions, and microscopic local details. Multi-level feature fusion maps texture features sequentially according to the triangular mesh hierarchy, allowing lower-level meshes to obtain overall texture information, while higher-level meshes obtain more refined local textures.

[0041] Constructing a 3D model with textured details based on the spatial coordinate relationships of multi-level feature fusion refers to reconstructing the final 3D solid model using the fused texture data and triangular mesh data as a foundation after feature fusion, through the coordinate relationships of spatial points, edges, and faces. Spatial coordinate relationships describe the 3D coordinate values ​​of vertices in the triangular mesh structure, mesh connectivity, and the projection position of textures within the geometric structure, ensuring the model remains spatially consistent with the actual construction object and presents realistic textures visually. The 3D model constructed from the fused data not only possesses high-precision geometric structure but also realistic texture details, enabling it to demonstrate greater expressive power in construction monitoring, progress analysis, and visualization.

[0042] Semantic segmentation based on 3D models, utilizing texture and point cloud structural features to identify different component objects, involves the joint analysis of texture features and point cloud geometric features on the model surface after geometric construction and texture bonding. This allows for automatic or semi-automatic classification and identification of the 3D model according to the boundaries of different engineering components. Texture features include information such as color gradients, surface textures, and material reflection characteristics, while point cloud structural features include data such as spatial curvature, boundary abrupt changes, geometric shape patterns, and triangular mesh topology. Semantic segmentation interprets these features semantically, inferring the boundaries between components through changes in texture and geometric boundaries within the model. For example, significant color differences and abrupt changes in point cloud normal vectors can distinguish between busbars, steel structure columns, or equipment foundations, enabling the model to be automatically split into multiple independent objects based on component type.

[0043] Each segmented component object is treated as an independent query entity in the database. Establishing structural data for each component means that after extracting the component object, it is stored as a separate data unit in a structured database, enabling it to be independently queried, accessed, and modified. Structural data refers to information related to the component, such as geometric dimensions, spatial coordinates, installation relationships, material properties, construction status, and coding identifiers. This data comprehensively characterizes the component's properties and supports subsequent construction management.

[0044] Adding structural data of each component to the 3D model refers to associating the structural data generated in the database with the corresponding component objects in the 3D model. This ensures that the 3D model not only contains geometric and textural information but also possesses complete attribute information, making it a queryable, analyzable, and traceable digital component model. By binding structural data to model objects, the 3D model can express information within an engineering scenario. For example, selecting a component in the model allows users to view its geometric dimensions, construction date, or inspection parameters, thereby achieving synchronous integration of the 3D model with engineering management data.

[0045] S3: Integrate the structural data based on the three-dimensional model, receive real-time construction monitoring data for status updates, and construct structured three-dimensional update data.

[0046] Furthermore, this application also includes: based on the component objects in the three-dimensional model, performing construction-related structural analysis, establishing a component construction structure model, and creating a unique digital identity code and associated attribute set for each component object; based on the component objects and the component construction structure model, receiving real-time construction monitoring data, inserting data according to the association mapping relationship of the component objects, updating the current construction spatial structural state of the component objects and the component-related attribute data, and constructing structured three-dimensional update data for the three-dimensional model.

[0047] Specifically, based on the component objects in the 3D model, construction association structure analysis is performed to establish a component construction structure model, and a unique digital identity code and associated attribute set are created for each component object. The 3D model is used to express the spatial geometric features and component topological relationships of the target project. By performing construction association structure analysis on the segmented component objects in the model, the semantic position of each component in the construction process, installation sequence, and upstream and downstream dependencies can be identified, thereby constructing a component construction structure model that can describe the entire process association logic of the component from processing, transportation, hoisting to positioning. In addition, the unique digital identity code generated for each component object is used to achieve unique identification in subsequent construction progress management, quality tracking, and data binding processes, while the component associated attribute set is used to record the component's dimensional parameters, material parameters, construction requirements, quality acceptance standards, and subsequent dynamically updated status information, thereby ensuring that a complete information mapping relationship can be established between each component object in the 3D model and the construction site.

[0048] Furthermore, based on component objects and component construction structure models, real-time construction monitoring data is received. Data is inserted according to the association mapping relationships of the component objects, updating the current spatial structural state of the component objects and their associated attribute data, thus constructing structured 3D update data for the 3D model. Component objects, as data organization units, work in conjunction with the construction process logic and spatial dependencies defined in the component construction structure model to classify and bind real-time monitoring data collected at the construction site. This real-time monitoring data includes LiDAR scan point clouds, image recognition progress markers, GNSS positioning results, vibration sensor signals, and progress messages output by the construction management system. Subsequently, according to the association mapping relationships of the component objects, the monitoring data is precisely inserted into the corresponding component nodes, reflecting the current geometric completion, installation offset, construction posture, and real-time construction environmental conditions, thereby enabling the component object's attribute set to be dynamically updated over time. This update process forms searchable, verifiable, and usable structured 3D update data for subsequent analysis, which is ultimately written back into the 3D model, giving it the ability to reflect the construction sequence and real-time status.

[0049] S4: Add the structured 3D update data to the 3D model, verify the parameters of the 3D map data, and fuse the verified 3D map parameters with the 3D model to generate a 3D map.

[0050] Furthermore, this application also includes: selecting at least N precise coordinate control points from the three-dimensional model, and installation constraints for the component object, wherein N is not less than 10; using the N precise coordinate control points to verify the coordinate accuracy of the three-dimensional map data, and obtaining coordinate verification results; using the installation constraints to verify the construction space and assembly parameters of the three-dimensional map data, and obtaining assembly verification results; when both the coordinate verification results and the assembly verification results pass, determining the three-dimensional map parameters and generating the three-dimensional map.

[0051] Furthermore, this application also includes: connecting to construction logs to obtain construction schedule plans, material information, quality inspection reports, construction monitoring parameters, and material traceability codes; performing structured data transformation on the obtained construction log data to generate construction record data, storing it in the corresponding component object data, and updating the structured 3D data of the 3D model.

[0052] Furthermore, this application also includes: obtaining a construction schedule plan; analyzing the construction schedule plan for construction parameters, construction target components, and construction objectives; simulating the construction parameters and construction target components in the three-dimensional map, searching for the optimal construction strategy to achieve the construction objectives, and generating a construction navigation path; integrating the construction navigation path into the three-dimensional map for construction visualization interaction, used for real-time construction navigation guidance.

[0053] Specifically, structured 3D update data is added to the 3D model. At least N precise coordinate control points are selected from the 3D model, along with installation constraints for the component objects, where N is no less than 10. Structured 3D update data refers to multi-dimensional, well-structured 3D spatial update information based on real-time monitoring data, component attribute sets, and construction status records. It reflects the dynamic changes of the component objects over time. Adding this data to the 3D model enables it to evolve dynamically. Subsequently, at least N precise coordinate control points are selected from the 3D model. These control points are spatial locations in the construction scene obtained by high-precision measuring instruments, possessing coordinate reference significance and usable for positioning calibration. When N is no less than 10, the coordinate calibration process can be guaranteed to meet statistical stability requirements. Simultaneously, the installation constraints for the component objects describe the component's posture, connection boundaries, geometric constraints, allowable deviation range, and other construction rules during spatial installation, providing a benchmark for subsequent verification.

[0054] The coordinate accuracy of the 3D map data is verified using N precise coordinate control points to obtain the verification results. Using the pre-selected N precise coordinate control points, their positions in the actual measurement coordinate system are compared point-by-point with the coordinates of their corresponding points in the 3D map data. By calculating indicators such as positional deviation, component error, and mean absolute deviation, the spatial coordinate accuracy of the 3D map data is verified to ensure it meets the accuracy requirements. The final coordinate verification results are used to determine whether the 3D model possesses engineering application capabilities at the coordinate level.

[0055] The installation constraints are used to verify the construction space and assembly parameters of the 3D map data, obtaining assembly verification results. The installation constraints provide the spatial layout specifications, structural connection requirements, and acceptable deviation ranges for component objects. Based on the geometric state, position, and spatial relationship of the current component in the 3D map data, each item is compared with the installation constraints to determine whether the construction space meets the clearance requirements for installation operations and whether the assembly parameters meet the design standards. As the number of components increases, the number of constraints for assembly parameter verification increases accordingly, but the structured 3D updated data ensures consistency and traceability throughout the verification process, ultimately forming the assembly verification results.

[0056] When both coordinate verification and assembly verification results pass, the 3D map parameters are determined, and a 3D map is generated. Based on the dual judgment of coordinate verification and assembly verification results, the reliability of the final 3D map parameters, including coordinate reference parameters, component geometric parameters, spatial attitude parameters, and relative relationship parameters between components, is only confirmed when both results meet the verification requirements. This generates a final 3D map that meets construction accuracy requirements and assembly logic consistency, which can be used for subsequent construction scheduling, assembly verification, digital twin display, and other application scenarios, ensuring that the 3D map has engineering-grade application accuracy.

[0057] By connecting to construction logs, we can obtain construction schedule plans, material information, quality inspection reports, construction monitoring parameters, and material traceability codes. Construction logs are collections of engineering management information recording the entire construction process, including multiple data types across time, space, and operational dimensions. By connecting to construction logs, we can obtain construction schedule plans (progress control information describing the expected start and end times of various construction tasks and the planned sequence of procedures); material information (material management information describing material numbers, specifications, supply batches, and inventory status); quality inspection reports (results such as structural quality, strength data, and pass / fail determinations obtained through quality inspections during construction); construction monitoring parameters (dynamic parameters such as displacement, settlement, vibration, and temperature acquired in real time through sensors and measuring equipment); and material traceability codes (unique identifiers identifying the source, flow process, and usage location of a material). This provides a data foundation for subsequent data structuring and 3D model updates.

[0058] The acquired construction log data undergoes structured data transformation to generate construction record data, which is then stored in the corresponding component object data. This updates the structured 3D data of the 3D model. Structured data transformation refers to converting text, image, tabular, or irregular format information from the original construction logs into structured data with fixed fields, clear attribute relationships, and computability through methods such as field extraction, format normalization, and attribute encoding. The generated construction record data describes the temporal behavior, inspection data, and material association information during the construction process. It is mapped to the unique identification code of each component object and stored in the corresponding component's data entry, forming a complete time-series construction record for each component object. The newly generated construction record data is injected into the structured 3D data of the 3D model. By updating component object attributes, construction status markers, and spatial location annotations, the 3D model maintains synchronization with the actual construction in the data dimension, achieving continuous dynamic evolution of the model.

[0059] Obtain the construction progress plan and analyze the construction parameters, construction target components, and construction targets of the construction progress plan. The construction progress plan is a time management document used to describe the execution sequence, time arrangement, resource allocation, and target process requirements of construction tasks. By obtaining this progress plan, construction parameters can be extracted, which are the control parameters affecting construction execution, such as the types of equipment, operation methods, personnel configurations, and process conditions required during construction; the construction target components refer to the specific component objects that are clearly required to be processed, installed, or inspected within a specific time period in the plan, which are used to guide the construction target positioning within the spatial scope; the construction target analysis is to decompose the logical relationships, preconditions, resource dependencies, and time windows of each task according to the progress plan, and can identify the specific requirements of each target component in different construction stages, thereby laying the input data foundation for subsequent construction simulation.

[0060] Perform construction simulation of the construction parameters and construction target components in the three-dimensional map, search for the best construction strategy to achieve the construction target, and generate a construction navigation path. Construction simulation refers to simulating the spatial layout, operation environment, equipment accessibility, and construction sequence of the construction site in the three-dimensional map based on construction parameters and construction target components. Through the dynamic calculation of the virtual construction environment, the impacts of different construction methods on progress, safety, spatial conflicts, and equipment efficiency are evaluated. The best construction strategy refers to an optimal path plan that comprehensively considers time cost, spatial feasibility, smoothness of equipment paths, and construction safety obtained through simulation search. Through simulation, the one with the fewest constraints can be selected to reduce problems such as equipment collisions and path blockages that may occur during construction. The finally generated construction navigation path is used to describe the movement trajectories of construction equipment and personnel in the three-dimensional space, including path nodes, attitude adjustment points, and key operation points.

[0061] Based on the construction navigation path, integrate it into the three-dimensional map for construction visualization interaction, which is used for real-time construction navigation guidance. Integrating the construction navigation path into the three-dimensional map means overlaying the optimized construction trajectory onto the spatial structure of the three-dimensional model, making the navigation path an integral part of the model, and enabling it to be linked and displayed with component objects and real-time monitoring data in the three-dimensional scene. Construction visualization interaction means that construction personnel can view the navigation path, target components, construction attitude, and equipment status in a three-dimensional manner through a graphical interface to form an intuitive operation guide. Construction real-time navigation guidance is used to dynamically prompt the next operation steps, the best travel route, and construction risks during the construction execution process based on the current position, construction progress, and on-site monitoring data, thereby improving construction efficiency and safety.

[0062] In summary, the three-dimensional map construction method for substation construction sites provided in this application has the following technical effects: by achieving the technical goal of digital modeling and dynamic updating of the substation construction site throughout its entire lifecycle, all components, and all elements, it achieves the technical effects of high-precision spatial restoration, real-time status perception, intelligent construction navigation guidance, and full-process visualized collaborative management in complex construction environments.

[0063] Example 2: Based on the same inventive concept as the three-dimensional map construction method for substation construction sites described in the foregoing examples, this application also provides a three-dimensional map construction system for substation construction sites. Please refer to the appendix. Figure 2 The system includes: a data processing module 1, used to collect point cloud data of the construction site using lidar, acquire corresponding image data using image acquisition equipment, perform filtering, downsampling, and coordinate registration processing on the point cloud data, and perform distortion correction and texture extraction on the image data; a fusion feature recognition module 2, used to align the point cloud data with the image data in coordinates, and recognize fusion features to generate a 3D model with texture details; a 3D update data construction module 3, used to integrate various structural data based on the 3D model, receive real-time construction monitoring data for status updates, and construct structured 3D update data; and a 3D map generation module 4, used to add the structured 3D update data to the 3D model, verify the parameters of the 3D map data, and fuse the verified 3D map parameters with the 3D model to generate a 3D map.

[0064] Furthermore, the 3D map construction system for substation construction sites is also used for: decomposing construction nodes based on substation construction to determine construction cycle nodes; locating construction objects in the construction area and construction targets according to the construction cycle nodes, and analyzing the acquisition perspectives based on the located construction objects to determine multiple acquisition perspectives, ensuring no blind spots in coverage, and setting high-precision targets as the common alignment benchmark for multi-cycle, multi-angle point cloud registration; configuring multi-period, multi-angle laser acquisition plans based on the construction cycle nodes and the corresponding multiple acquisition perspectives, setting corresponding high-precision targets, and acquiring point cloud data of the construction site.

[0065] Furthermore, the 3D map construction system for substation construction sites is also used for: aligning point cloud data with image data according to the acquisition target and coordinates; generating a multi-level triangular mesh model from macro to micro based on the original high-density point cloud; matching the texture source of visual angles using the normal vectors of the point cloud data and the image data based on the coordinate alignment relationship, and establishing a texture alignment relationship between the point cloud data and the image data; adding the texture alignment relationship to the multi-level triangular mesh model to establish texture coordinate alignment between the image data texture and the multi-level triangular network.

[0066] Furthermore, the 3D map construction system for substation construction sites is also used for: performing multi-level feature fusion based on the texture coordinate alignment relationship between the image data texture and the multi-level triangular network, wherein the texture features of the image data are fused with the hierarchical structure features of the point cloud data; and constructing the 3D model with texture details according to the spatial coordinate relationship of the multi-level feature fusion.

[0067] Furthermore, the 3D map construction system for substation construction sites is also used for: semantic segmentation of texture features and point cloud structure features based on the 3D model to determine different component objects; establishing the structural data of each component as an independent query entity in the database; and adding the structural data of each component to the 3D model.

[0068] Furthermore, the 3D map construction system for substation construction sites is also used for: analyzing the construction-related structures based on the component objects in the 3D model, establishing a component construction structure model, and creating a unique digital identity code and associated attribute set for each component object; receiving real-time construction monitoring data based on the component objects and the component construction structure model, inserting data according to the association mapping relationship of the component objects, updating the current construction spatial structure status of the component objects and the component association attribute data, and constructing structured 3D update data for the 3D model.

[0069] Furthermore, the 3D map construction system for substation construction sites is also used for: selecting at least N precise coordinate control points and installation constraints of component objects from the 3D model, wherein N is not less than 10; verifying the coordinate accuracy of the 3D map data using the N precise coordinate control points to obtain coordinate verification results; verifying the construction space and assembly parameters of the 3D map data using the installation constraints to obtain assembly verification results; and determining the 3D map parameters and generating the 3D map when both the coordinate verification results and the assembly verification results pass.

[0070] Furthermore, the 3D map construction system for substation construction sites is also used to: connect to construction logs to obtain construction progress plans, material information, quality inspection reports, construction monitoring parameters, and material traceability codes; perform structured data transformation on the obtained construction log data to generate construction record data, store it in the corresponding component object data, and update the structured 3D data of the 3D model.

[0071] Furthermore, the 3D map construction system for substation construction sites is also used for: acquiring a construction schedule plan; analyzing the construction schedule plan for construction parameters, construction target components, and construction objectives; simulating the construction parameters and construction target components in the 3D map; searching for the optimal construction strategy to achieve the construction objectives; and generating a construction navigation path; integrating the construction navigation path into the 3D map for interactive construction visualization, used for real-time construction navigation guidance.

[0072] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The three-dimensional map construction method and specific examples for substation construction sites described in Embodiment 1 are also applicable to the three-dimensional map construction system for substation construction sites in this embodiment. Through the foregoing detailed description of the three-dimensional map construction method for substation construction sites, those skilled in the art can clearly understand the three-dimensional map construction system for substation construction sites in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0073] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for constructing a three-dimensional map for a substation construction site, characterized in that, include: Point cloud data of the construction site is collected by lidar, and corresponding image data is obtained by image acquisition equipment. The point cloud data is filtered, downsampled and coordinate registered, while the image data is subjected to distortion correction and texture extraction. Aligning the point cloud data with the image data in coordinates and identifying fusion features to generate a 3D model with texture details includes: Based on the 3D model, semantic segmentation of texture features and point cloud structure features is performed to determine different component objects; Each segmented component object is treated as an independent query entity in the database, and the structural data of each component is established. Add the structural data of each component to the three-dimensional model; Based on the aforementioned 3D model, structural data are integrated, and real-time construction monitoring data is received for status updates to construct structured 3D updated data, including: Based on the component objects in the three-dimensional model, the construction association structure is analyzed, a component construction structure model is established, and a unique digital identity code and associated attribute set are created for each component object. Based on the component object and the component construction structure model, real-time construction monitoring data is received, data is inserted according to the association mapping relationship of the component object, the current construction spatial structure status of the component object and the component association attribute data are updated, and structured three-dimensional update data of the three-dimensional model is constructed. The structured 3D update data is added to the 3D model, the parameters of the 3D map data are verified, and the verified 3D map parameters are fused with the 3D model to generate a 3D map.

2. The method for constructing a three-dimensional map of a substation construction site according to claim 1, characterized in that, Point cloud data of the construction site is collected using lidar, including: Based on the substation construction, construction nodes are decomposed to determine the construction cycle nodes; Based on the construction cycle nodes, the construction objects of the construction area and construction targets are located, and the collection perspective is analyzed based on the located construction objects to determine multiple collection perspectives to ensure coverage without blind spots. High-precision targets are set and used as the common alignment benchmark for multi-cycle and multi-angle point cloud registration. Based on the construction cycle nodes and the corresponding multi-view acquisition angles, a multi-phase, multi-view laser acquisition plan is configured, and corresponding high-precision targets are set to acquire point cloud data of the construction site.

3. The method for constructing a three-dimensional map of a substation construction site according to claim 1, characterized in that, Aligning the point cloud data with the image data in terms of coordinates includes: Align the point cloud data with the image data according to the acquisition target and acquisition coordinates; Based on the original high-density point cloud, a multi-level triangular mesh model from macroscopic to microscopic is generated; Based on the coordinate alignment relationship, the texture source matching of visual angle is performed by using the normal vector of point cloud data and image data to establish the texture alignment relationship between point cloud data and image data. Texture alignment relationships are added to the multi-level triangular mesh model to establish the alignment of image data texture with the texture coordinates of the multi-level triangular network.

4. The method for constructing a three-dimensional map of a substation construction site according to claim 3, characterized in that, Identify and fuse features to generate a 3D model with texture details, including: Based on the alignment relationship between the texture of the image data and the texture coordinates of the multi-level triangular network, multi-level feature fusion is performed, wherein the texture features of the image data are fused with the hierarchical structure features of the point cloud data. Based on the spatial coordinate relationship of the multi-level feature fusion, the 3D model with texture details is constructed.

5. The method for constructing a three-dimensional map of a substation construction site according to claim 1, characterized in that, The parameters of the 3D map data are validated, and the validated 3D map parameters are fused with the 3D model to generate a 3D map, including: Select at least N precise coordinate control points from the three-dimensional model, as well as installation constraints for the component object, where N is not less than 10; The coordinate accuracy of the three-dimensional map data is verified using the N precise coordinate control points to obtain the coordinate verification results. The installation constraints are used to verify the construction space and assembly parameters of the three-dimensional map data to obtain the assembly verification results. When both the coordinate verification result and the assembly verification result pass, the 3D map parameters are determined and the 3D map is generated.

6. The method for constructing a three-dimensional map of a substation construction site according to claim 5, characterized in that, The structured 3D update data for constructing the 3D model also includes: Connect to construction logs to obtain construction schedule plans, material information, quality inspection reports, construction monitoring parameters, and material traceability codes; The acquired construction log data is transformed into structured data to generate construction record data, which is then stored in the corresponding component object data. The structured 3D data of the 3D model is then updated.

7. The method for constructing a three-dimensional map of a substation construction site according to claim 6, characterized in that, After generating the 3D map, the following is also included: Obtain the construction schedule plan and analyze the construction parameters, target components, and objectives of the construction schedule plan. The construction parameters and target components are simulated on the 3D map to search for the best construction strategy to achieve the construction objective and generate a construction navigation path. The construction navigation path is integrated into the 3D map to enable interactive construction visualization and provide real-time navigation guidance.

8. A three-dimensional map construction system for substation construction sites, characterized in that, The steps for implementing the three-dimensional map construction method for a substation construction site according to any one of claims 1 to 7 include: The data processing module is used to collect point cloud data of the construction site through lidar, obtain corresponding image data through image acquisition equipment, filter, downsample and coordinate registration of point cloud data, and perform distortion correction and texture extraction on image data. A fusion feature recognition module is used to align the point cloud data with the image data in coordinates, and to recognize fusion features to generate a 3D model with texture details, including: Based on the 3D model, semantic segmentation of texture features and point cloud structure features is performed to determine different component objects; Each segmented component object is treated as an independent query entity in the database, and the structural data of each component is established. Add the structural data of each component to the three-dimensional model; A three-dimensional update data construction module is used to integrate various structural data based on the three-dimensional model, receive real-time construction monitoring data for status updates, and construct structured three-dimensional update data, including: Based on the component objects in the three-dimensional model, the construction association structure is analyzed, a component construction structure model is established, and a unique digital identity code and associated attribute set are created for each component object. Based on the component object and the component construction structure model, real-time construction monitoring data is received, data is inserted according to the association mapping relationship of the component object, the current construction spatial structure status of the component object and the component association attribute data are updated, and structured three-dimensional update data of the three-dimensional model is constructed. The 3D map generation module is used to add the structured 3D update data to the 3D model, verify the parameters of the 3D map data, and fuse the verified 3D map parameters with the 3D model to generate a 3D map.

Images