Highway bridge disease data structured acquisition and tracking inspection method

Abstract

The invention provides a highway bridge disease data structured acquisition and tracking inspection method, and relates to the technical field of data structure establishment, and the method comprises the following steps: constructing a highway bridge structure analysis standard and disease metadata system; performing component coding on the disease classification result based on a preset evaluation standard, and dividing risk levels by evaluating disease degrees to obtain spatial position data; constructing a coordinate conversion system for disease space positioning; the bridge disease model and the bridge structure analysis model are fused, and a bridge defect visualization model used for presenting the disease position is generated; and in combination with historical disease data, carrying out tracking and comparative analysis on a disease development state, evaluating a disease development trend through the prediction model, and generating a bridge maintenance suggestion and a repair scheme. According to the invention, full-process digital management and control of diseases from collection, classification and positioning to tracking and maintenance are realized, and a solid technical support is provided for safety guarantee and long-term operation and maintenance of a bridge structure.

Description

Technical Field

[0001] This invention relates to the field of data structure technology, and more specifically, to a method for structured collection and tracking inspection of highway bridge defects data. Background Technology

[0002] As a critical node in transportation infrastructure, the structural safety and long-term operation of highway bridges are of paramount importance. Traditional bridge defect inspection mainly relies on manual visual inspection and simple tool measurement. Defect information is mostly described in natural language and recorded in paper or electronic spreadsheets with photographs. This method has inherent defects such as vague defect descriptions, inaccurate spatial positioning, and low data standardization, making it difficult for computers to accurately understand and efficiently utilize the defect information.

[0003] Currently, with the development of digital technologies such as 3D laser scanning, photogrammetry, and BIM, some research has begun to explore the application of these new technologies to bridge inspection. For example, point cloud data is used for surface deformation analysis, or BIM models are used for asset information management. However, existing solutions often have the following limitations: establishing a bridge component coding system, defect classification and description standards, and coordinate positioning standards that span the entire process of data acquisition, processing, analysis, and application leads to poor data consistency, making it difficult to achieve automated processing and cross-period, cross-project data comparison and accumulation. Furthermore, most methods remain at the stage of defect discovery and recording, failing to effectively utilize historical data for tracking defect development, trend prediction, and intelligent maintenance decision support based on prediction results.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] In view of this, the present invention provides a method for structured collection and tracking inspection of highway bridge defects data to solve the aforementioned problems.

[0006] To solve the above problems, the specific technical solution adopted by the present invention is as follows:

[0007] According to a first aspect of the present invention, a method for structured acquisition and tracking inspection of highway bridge defects is provided, the method comprising the following steps:

[0008] S1. Based on the pre-constructed bridge structure analytical model using BIM technology, perform component analysis on the pre-acquired highway bridge drawings, materials, and on-site measured data, and construct a highway bridge structure analytical standard and defect metadata system based on the component analysis results.

[0009] S2. Based on the highway bridge structure analysis standard and the disease metadata system, collect highway bridge disease data and classify the disease data; based on the preset evaluation standard, encode the disease classification results into components, and classify the risk level by assessing the degree of disease, to obtain spatial location data containing disease component codes, disease types, and risk levels.

[0010] S3. Based on the spatial location data, construct a bridge defect model, and combine it with the bridge structural analytical model to construct a coordinate transformation system for spatial positioning of defects;

[0011] S4. Based on the coordinate transformation system, the bridge defect model and the bridge structural analytical model are integrated to generate a bridge defect visualization model to present the location of defects.

[0012] S5. Based on the bridge defect visualization model and combined with historical defect data, track and compare the development status of defects, and evaluate the development trend of defects through the prediction model to generate bridge maintenance suggestions and repair plans.

[0013] Preferably, the process of constructing a bridge structural analytical model in advance using BIM technology, performing component analysis on pre-acquired highway bridge drawings, materials, and on-site measured data, and constructing a highway bridge structural analytical standard and defect metadata system based on the component analysis results includes the following steps:

[0014] S11. Obtain highway bridge drawings and materials and on-site measurement data, and based on the highway bridge drawings and materials, use BIM software to construct an analytical model of the bridge structure;

[0015] S12. Perform component analysis on the analytical model of the highway bridge structure, and determine the origin coordinate system of the component surface by extracting component features, forming a construction analysis result that includes component coding rules and component hierarchical relationships;

[0016] S13. Based on the component analysis results, configure the bridge structure analysis standard, which includes component coding rules, origin coordinate system definition, component hierarchical relationship and spatial position requirements;

[0017] S14. Based on the predefined disease classification and severity assessment rules, generate a disease metadata system that includes standard disease type codes, disease attribute field definitions, quantitative grading thresholds, and spatial data record specifications associated with component codes and the origin coordinate system.

[0018] Preferably, the process of collecting highway bridge defect data based on highway bridge structural analysis standards and defect metadata system, classifying the defect data, encoding the defect classification results into components based on preset evaluation standards, and classifying risk levels by assessing the degree of defect, to obtain spatial location data containing defect component codes, defect types, and risk levels includes the following steps:

[0019] S21. Use data acquisition equipment to acquire point cloud data and image data of the surface of highway bridges;

[0020] S22. Preprocess the image data using an image recognition algorithm to extract disease feature parameters; and verify the disease area through point cloud curvature analysis to obtain a set of disease feature parameters containing disease spatial coordinates and disease geometric features.

[0021] S23. Based on the set of defect feature parameters and combined with the bridge structure analysis standard, use the spatial inclusion relationship algorithm to determine the defect-related components, and obtain the defect component code based on the defect-related components;

[0022] S24. Based on the disease metadata system, the support vector machine algorithm is used to classify the disease feature parameter set to obtain the disease type;

[0023] S25. Based on the disease type and disease characteristic parameter set, calculate the disease risk level according to the quantitative grading threshold in the disease metadata system;

[0024] S26. The defects of the components, the types of defects, and the risk levels are encapsulated in a structured manner to generate spatial location data of highway bridge defects.

[0025] Preferably, the step of determining the associated components of a defect based on a set of defect feature parameters and in conjunction with bridge structure analysis standards, using a spatial inclusion relationship algorithm, and obtaining the defect component code based on the associated components includes the following steps:

[0026] S231. Based on the spatial coordinates of the defects in the defect feature parameter set, and combined with the bridge structure analysis standard, the point cloud coordinates of the defect area are transformed to the coordinate system of the bridge structure analysis model to obtain the transformed spatial coordinates of the defects.

[0027] S232. Using a pre-built spatial index structure and combined with the converted spatial coordinates of the disease, perform component queries to initially screen out the set of components whose spatial boundaries intersect with the disease area; and calculate the spatial relationship matching cost between the initially screened components and the disease area based on spatial relationship constraints, sort the components according to the matching cost, and select several components as a candidate component set.

[0028] S233. For each candidate component, the point cloud surface matching algorithm is used to determine whether the point cloud coordinates of the diseased area are located on the surface of the component, and the overlap ratio between the point cloud of the diseased area and the surface of the component is generated based on the determination result.

[0029] S234. Calculate the overlap ratio between the point cloud of the diseased area and the surface of each candidate component, and select the component with the highest overlap ratio that exceeds the preset threshold as the disease-related component.

[0030] S235. Based on the disease-related components and the geometric features in the disease characteristic parameter set, determine the final disease component code.

[0031] Preferably, the step of using a pre-built spatial index structure, combined with the converted spatial coordinates of the disease, to query components and initially filter out a set of components whose spatial boundaries intersect with the disease area; and calculating the spatial relationship matching cost between the initially filtered components and the disease area based on spatial relationship constraints, sorting the components according to the matching cost, and selecting several components as a candidate component set includes the following steps:

[0032] S2321. Based on the transformed spatial coordinates of the disease, a pre-built spatial index structure is used to query components, and all components intersecting with the disease area are determined by calculating the spatial bounding box, forming a preliminary set of selected components.

[0033] S2322. For each component in the preliminary selected component set, extract its spatial relationship feature vector with the diseased area;

[0034] S2323. Based on the spatial relationship feature vector, use the preset matching cost function to calculate the spatial relationship matching cost between each component and the diseased area;

[0035] S2324. Based on the calculated spatial relationship matching cost, sort the components in the preliminary screening component set in ascending order, and select several components with the smallest matching cost as the candidate component set.

[0036] Preferably, the step of querying components based on the transformed spatial coordinates of the disease, using a pre-built spatial index structure, and determining all components intersecting with the disease area by calculating the spatial bounding box to form a preliminary set of selected components includes the following steps:

[0037] S23211. For each component in the analytical model of the bridge structure, calculate the minimum and maximum coordinate points of its spatial bounding box and store them in a pre-built spatial index structure.

[0038] S23212. Convert the transformed set of spatial coordinates of the disease into a spatial bounding box of the disease area, and calculate the minimum and maximum coordinate points of the disease point cloud coordinates.

[0039] S23213. Using a pre-built spatial index structure, the spatial bounding box intersection determination algorithm is used to query all components that intersect with the bounding box of the diseased area to obtain the query results.

[0040] S23214. Extract component codes from the query results and generate a preliminary set of filtered components containing the component codes and spatial bounding box coordinates of all intersecting components.

[0041] Preferably, for each candidate component, determining whether the point cloud coordinates of the defective area are located on the component surface using a point cloud surface matching algorithm, and generating the overlap ratio between the point cloud of the defective area and the component surface based on the determination result, includes the following steps:

[0042] S2331. For each candidate component, based on the transformed spatial coordinates of the disease and the surface data of the component, calculate the shortest spatial distance from each point in the point cloud of the diseased area to the surface of the component.

[0043] S2332. Based on the preset distance tolerance threshold, points whose shortest spatial distance is less than or equal to the distance tolerance threshold are identified as defect points that overlap with the surface of the component.

[0044] S2333. Count the number of disease points that are judged to be overlapping, and calculate the proportion of them to the total number of points in the disease area point cloud, as the overlap ratio between the candidate component and the disease area point cloud.

[0045] According to a second aspect of the present invention, a system for structured acquisition and tracking inspection of highway bridge defects is provided, the system comprising:

[0046] The analysis module is used to analyze the components of the pre-built bridge structure analysis model using BIM technology, and to construct a highway bridge structure analysis standard and a defect metadata system based on the pre-acquired highway bridge drawings, materials and on-site measured data.

[0047] The disease analysis module is used to collect highway bridge disease data based on the highway bridge structure analysis standard and disease metadata system, and classify the disease data; based on the preset evaluation standard, the disease classification results are coded into components, and the risk level is divided by assessing the degree of disease, resulting in spatial location data containing disease component codes, disease types, and risk levels.

[0048] The coordinate transformation construction module is used to build a bridge defect model based on spatial location data, and combine it with the bridge structural analytical model to build a coordinate transformation system for spatial positioning of defects;

[0049] The visualization model generation module is used to fuse the bridge defect model with the bridge structure analytical model based on the coordinate transformation system, and generate a visualization model of bridge defects to show the location of defects.

[0050] The bridge damage prediction and assessment module is used to track and compare the development status of bridge damage based on a bridge damage visualization model and historical damage data. It also uses the prediction model to assess the development trend of damage and generate bridge maintenance suggestions and repair plans.

[0051] According to a third aspect of the present invention, an electronic device is provided, the electronic device comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the programs to implement the steps of the above-described method.

[0052] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein a computer program is stored therein, wherein the steps of the above-described method are implemented when the computer program controls the device in which the computer-readable storage medium is located to execute during runtime.

[0053] The beneficial effects of this invention are as follows:

[0054] 1. This invention achieves structured and standardized management of bridge defect data through bridge structure analysis standards and a defect metadata system; by leveraging multi-source data acquisition, precise coordinate transformation, and model fusion technology, it enables precise location and visualization of defect positions, significantly improving the accuracy and efficiency of defect detection; through historical data comparison and analysis and predictive model application, it can scientifically track the development status of defects and predict trends, providing differentiated and targeted suggestions and repair solutions for bridge maintenance, avoiding blind maintenance, reducing maintenance costs, and simultaneously realizing full-process digital control of defects from collection, classification, location to tracking and maintenance, providing solid technical support for bridge structural safety and long-term operation and maintenance.

[0055] 2. This invention is based on multi-source data acquisition of point clouds and images, combined with image recognition algorithms and point cloud curvature analysis to achieve accurate extraction of disease features and removal of false diseases, ensuring data authenticity. Through a hierarchical screening mechanism of pre-constructed spatial index, spatial bounding box intersection determination, spatial relationship matching cost ranking, and point cloud surface overlap ratio verification, it accurately locates disease-related components and obtains unique component codes. With the help of support vector machine algorithms and the quantitative threshold of the disease metadata system, it achieves accurate classification of disease types and scientific determination of risk levels, ensuring the standardization and consistency of classification and grading. Finally, through structured encapsulation, it generates complete data containing component codes, types, risk levels, and spatial coordinates, which not only ensures the standardization, relevance, and traceability of disease data, but also significantly improves the efficiency and accuracy of disease acquisition and processing, providing a high-quality, standardized core data source for subsequent bridge disease model construction, spatial positioning, and maintenance decisions. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0057] Figure 1 This is a flowchart of a method for structured data collection and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of the bridge structure analysis hierarchy in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0059] Figure 3 This is a schematic diagram of bridge structure analysis in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0060] Figure 4 This is a schematic diagram of a bridge defect model in a method for structured data collection and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0061] Figure 5 This is a schematic diagram of bridge components and component surfaces in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0062] Figure 6 This is a schematic diagram illustrating the defined origin position of each component surface in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0063] Figure 7 This is a schematic diagram illustrating the reproduction of bridge defects in a method for structured data collection and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0064] Figure 8 This is a schematic diagram of the details of the defect model in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0065] Figure 9 This is a schematic diagram of a system for structured data acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0066] Figure 10 This is a schematic diagram of the defect markers, defect attributes, and unique codes in a method for structured collection and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0067] Figure 11 This is a schematic diagram of the defect attributes in a method for structured collection and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0068] Figure 12 This is a schematic diagram of component attributes in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0069] Figure 13 This is a schematic diagram illustrating the import of parameter data for multi-column piers in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0070] Figure 14 This is one of the schematic diagrams of a defect model in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0071] Figure 15 This is the second schematic diagram of a defect model in a method for structured data collection and tracking inspection of highway bridge defects according to an embodiment of the present invention;

[0072] Figure 16 This is a hardware structure block diagram of the host device in a method for structured acquisition and tracking inspection of highway bridge defects according to an embodiment of the present invention.

[0073] In the picture:

[0074] 1. Construction and analysis module; 2. Disease analysis module; 3. Coordinate transformation construction module; 4. Visualization model generation module; 5. Disease prediction and assessment module; 6. Top plate; 7. Inner web plate; 8. Cavity; 9. Bottom plate; 10. Outer web plate; 11. Outer wing plate; 12. Inner wing plate inner side surface; 13. Inner web plate inner side surface one; 14. Inner web plate inner side surface two; 15. Bottom plate inner side surface; 16. Bottom plate outer side surface; 17. Outer web plate outer side surface; 18. Outer wing plate outer side surface; 19. Top plate outer side surface. Detailed Implementation

[0075] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0076] The methods and embodiments provided in this application can be executed on a host device or a similar computing device. Taking running on a host device as an example, such as... Figure 16 As shown, the host device may include one or more ( Figure 16 Only one is shown in the diagram. The processor (which may include, but is not limited to, a microprocessor (MCU) or programmable logic device (FPGA), etc.) and storage for storing data are also shown. The host device may further include transmission devices for communication functions and input / output devices. Those skilled in the art will understand that... Figure 16 The structure shown is for illustrative purposes only and does not limit the structure of the host device described above. For example, the host device may also include components that are larger than... Figure 16 The more or fewer components shown, or having the same Figure 16 The different configurations shown.

[0077] The memory can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the exception handling method in this embodiment. The processor executes various functional applications and data processing by running the computer program stored in the memory, thus implementing the above-described method. The memory may include high-speed random access memory (RAM) and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the host device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks (LANs), mobile communication networks, and combinations thereof.

[0078] Transmission devices are used to receive or send data over a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the host device. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0079] According to an embodiment of the present invention, a method for structured acquisition and tracking inspection of highway bridge defects is provided.

[0080] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to a first embodiment of the present invention, a method for structured acquisition and tracking inspection of highway bridge defects is provided, the method comprising the following steps:

[0081] S1. Based on the pre-constructed bridge structure analytical model using BIM technology, perform component analysis on the pre-acquired highway bridge drawings, materials, and on-site measured data, and construct a highway bridge structure analytical standard and defect metadata system based on the component analysis results.

[0082] As a preferred embodiment, the process of constructing a bridge structural analytical model in advance using BIM technology, analyzing pre-acquired highway bridge drawings, materials, and on-site measured data, and constructing a highway bridge structural analytical standard and defect metadata system based on the component analysis results includes the following steps:

[0083] S11. Obtain highway bridge drawings and materials and on-site measurement data, and based on the highway bridge drawings and materials, use BIM software to construct an analytical model of the bridge structure;

[0084] It should be noted that highway bridge drawings include CAD construction drawings, structural calculation sheets, component lists, etc. The structural calculation sheets contain data such as component materials, dimensions, and stress parameters, while the component lists include component numbers, quantities, and installation locations. The actual on-site measurement data is obtained by using a laser scanner to acquire 3D point cloud data of the bridge surface, with a point cloud density ≥100 points / cm² during acquisition. 2 The accuracy is ±2mm. Global coordinates of control points for key components were collected using GPS / RTK equipment, and overall bridge appearance data was obtained using UAV oblique photography to verify the overall shape of the model.

[0085] The BIM software used is Revit. By importing DWG format files from CAD construction drawings, the software's drawing vectorization and component parametric modeling functions convert the two-dimensional drawings into three-dimensional parametric models. Then, the core attributes of the components are entered according to the drawing parameters, including material, cross-sectional dimensions, installation location, angles, etc. Figure 13 The image shows the parameter data for the multi-column pier, including angle and installation location information, thus enabling the creation of an initial analytical model of the bridge structure. The laser-scanned point cloud data is then imported into BIM software and registered and aligned with the initial analytical model. A deviation analysis tool is used to compare the geometric differences between the model and the point cloud, such as discrepancies between the actual dimensions of components and the drawings, and offsets in installation location. Based on these differences, the model parameters are adjusted to ultimately form the analytical model of the bridge structure. Figure 3 As shown.

[0086] S12. Perform component analysis on the analytical model of the highway bridge structure, and determine the origin coordinate system of the component surface by extracting component features, forming a construction analysis result that includes component coding rules and component hierarchical relationships;

[0087] It should be noted that, as Figure 2 As shown, to achieve precise assessment and facilitate accurate maintenance, in-depth analysis and coding of bridge structures are necessary. Based on the four-layer bridge structural analysis (bridge > location > component > member) established in the *Highway Bridge Technical Condition Assessment Standard* (JTG / TH21-2011), this is expanded upwards and downwards to form seven levels: "bridge - assessment unit - location - member - sub-member - member surface." This constitutes a seven-layer analysis rule that satisfies the three-system (beam, arch, cable) bridge structure of expressways. The bridge structural hierarchy is divided into seven levels: bridge, assessment unit, location, component, member, sub-member, and member surface. Locations include the superstructure, substructure, bridge deck system, and ancillary facilities. Components include superstructure load-bearing members, superstructure general members, piers, abutments, expansion joint devices, maintenance facilities, anti-throw nets, etc. The classification of bridge assessment units, locations, and members is consistent with the requirements of the current *Highway Bridge Technical Condition Assessment Standard* (JTG / TH21) and *Highway Bridge and Culvert Maintenance Specification* (JTG5120). The structural objects at each level of the bridge should be expandable. Newly expanded objects should belong to the structural levels specified in this document, and should include cast-in-place box girders, diaphragms, column piers, U-shaped abutments, asphalt concrete, steel expansion joints, maintenance vehicles, rectangular anti-throw nets, etc. Sub-components are different functional parts that make up the bridge components, such as anchorages in a cable-stayed system, stop blocks in a cap beam, and toothed plates in a cast-in-place box girder. Component surfaces are two-dimensional surfaces of components or sub-components, representing the lowest level of bridge component division, such as the bottom plate and outer web of the main girder, and the side of the cap beam at its largest station.

[0088] Specifically, by structurally decomposing the BIM model, core features are extracted to form a feature library, including: geometric features, attribute features, and location features; geometric features include dimensional parameters and cross-sectional shape; attribute features include material type, design load level, and stress type; location features include global coordinate range, associated station number, and connection relationship with adjacent components.

[0089] Among them, such as Figure 5 As shown, the components include a top plate 6, an inner web plate 7, a cavity plate 8, a bottom plate 9, an outer web plate 10, and an outer wing plate 11, etc. When determining the origin coordinate system of the component surface, based on the geometric characteristics of the component surface and the operability of engineering inspection, a stable, unique, and easily identifiable physical feature point on-site is defined as the origin of the two-dimensional local coordinate system of that surface, such as... Figure 6 As shown, this involves the inner surface of the inner flange 12, the inner surface of the inner web 13, the inner surface of the inner web 2 14, the inner surface of the bottom plate 15, the outer surface of the bottom plate 16, the outer surface of the outer web 17, the outer surface of the outer flange 18, and the outer surface of the top plate 19. For example, for the bottom surface of a rectangular beam, a corner point is selected. For the surface of a cylindrical pier, the intersection of its bottom circumference and a generatrix in a specific direction is selected. For an inclined web surface, a specific endpoint on the intersection line with the adjacent bottom or top surface can be selected. Based on this origin, two mutually perpendicular coordinate axes are defined according to the main direction of the component. Typically, the U-axis is along the main longitudinal or circumferential direction of the component, and the V-axis is along the vertical or transverse direction. The origin coordinates and axial directions of this local coordinate system are calculated and recorded. Then, according to the pre-designed coding rules, a globally unique identification code is generated for each component surface and its associated component unit. Finally, the feature records of all component units, the definition of the origin coordinate system of each component surface, and its hierarchical relationship are integrated and output into a structured construction analysis result file or database. The pre-designed coding rules include route codes, bridge codes, component type codes, unit serial numbers, surface type codes, etc.

[0090] S13. Based on the component analysis results, configure the bridge structure analysis standard, which includes component coding rules, origin coordinate system definition, component hierarchical relationship and spatial position requirements;

[0091] It should be noted that, based on the component analysis results, the seven-level hierarchical division rule of "bridge-evaluation unit-location-component-component-subcomponent-component surface" is standardized and solidified. The globally unique coding rules composed of route code, bridge code, component type code, unit number, and surface type code, as well as the assignment logic of each level of coding, are clearly defined. The selection criteria for the origin of each type of component surface, the definition rules of the U-axis and V-axis of the local coordinate system, and the bidirectional transformation parameters between the local coordinate system and the global geodetic coordinate system are unified. At the same time, the spatial position accuracy requirements are quantified, and it is clarified that the component coordinate deviation and origin positioning deviation shall not exceed 3mm, and the geometric matching deviation between the model and the laser point cloud shall be ≤±2mm. Finally, a bridge structure analysis standard covering component coding rules, origin coordinate system definition, component hierarchical relationship, and spatial position accuracy requirements is formed.

[0092] S14. Based on the predefined disease classification and severity assessment rules, generate a disease metadata system that includes standard disease type codes, disease attribute field definitions, quantitative grading thresholds, and spatial data record specifications associated with component codes and the origin coordinate system.

[0093] It should be noted that, based on the "Standard for Technical Condition Assessment of Highway Bridges" (JTG / TH21-2011) and the "Specifications for Maintenance of Highway Bridges and Culverts" (JTG5120), and referring to predefined rules for disease classification and severity assessment, a three-level disease classification system (major category, intermediate category, minor category) was constructed, and corresponding standard disease type codes were generated. Structured attribute fields containing basic disease information, morphological parameters, and spatial information were designed, such as... Figure 11 As shown; then, combining the material type, stress characteristics and design load level of different components, we formulate differentiated quantitative classification thresholds for defects and determine the judgment criteria for four levels of defects: minor, moderate, relatively severe and severe; at the same time, we formulate spatial data recording specifications that are directly related to component codes and origin coordinate systems, set the recording format, accuracy requirements and association logic for local and global coordinates of defects, and finally generate a complete defect metadata system.

[0094] S2. Based on the highway bridge structure analysis standard and the disease metadata system, collect highway bridge disease data and classify the disease data; based on the preset evaluation standard, encode the disease classification results into components, and classify the risk level by assessing the degree of disease, to obtain spatial location data containing disease component codes, disease types, and risk levels.

[0095] As a preferred embodiment, the process of collecting highway bridge defect data based on highway bridge structural analysis standards and defect metadata system, classifying the defect data, encoding the defect classification results into components based on preset evaluation standards, and classifying risk levels by assessing the degree of defect, to obtain spatial location data containing defect component codes, defect types, and risk levels includes the following steps:

[0096] S21. Use data acquisition equipment to acquire point cloud data and image data of the surface of highway bridges;

[0097] Specifically, the data acquisition equipment includes laser scanners, high-definition industrial cameras, and GPS / RTK devices. The acquisition process requires targeting the corresponding component units and surfaces in the bridge structure analysis standards. The laser scanner is used to acquire three-dimensional point cloud data of the bridge surface, with a point cloud density of no less than 100 points / cm² during acquisition. 2 The spatial accuracy is controlled within ±2mm, accurately restoring the surface morphology of components and the three-dimensional location of defects. High-definition industrial cameras with at least 12 megapixels are required to capture panoramic, close-up, and extreme close-up images of the defect area, clearly recording the visual characteristics of the defects. GPS / RTK equipment simultaneously acquires the global geodetic coordinates of key component control points and the center of the defect area. After acquisition, all data is associated with the corresponding component units using temporary identifiers.

[0098] S22. Preprocess the image data using an image recognition algorithm to extract disease feature parameters; and verify the disease area through point cloud curvature analysis to obtain a set of disease feature parameters containing disease spatial coordinates and disease geometric features.

[0099] Specifically, the image preprocessing stage uses the YOLOv8 image recognition algorithm. This algorithm can adapt to the complex scenes of bridge defects and can effectively resist changes in lighting and shadow interference. Through image denoising, enhancement, segmentation and other operations, it extracts feature parameters of different types of defects, including: for crack defects, extract parameters such as length, maximum width, depth and direction; for surface defect defects, extract parameters such as area, depth and corrosion level; and for displacement defects, extract parameters such as offset and tilt.

[0100] In point cloud curvature analysis, the curvature values ​​of each point in the point cloud data can be calculated, and the curvature threshold set by the bridge structure analysis standard can be used. Specifically, based on the bridge structure analysis standard, the design parameters and structural characteristics of the corresponding bridge type and components are extracted. For healthy components without defects, the theoretical curvature range of their design surfaces is extracted through the BIM structural analysis model. For example, the theoretical curvature of the web of a flat box girder is close to 0, and the curved arch rib has a fixed design curvature value. Combined with the "Highway Bridge Technical Condition Assessment Standard", the upper limit of curvature in the healthy area and the lower limit of curvature in the defective area are determined. For example, the curvature of the point cloud on the surface of a healthy component should be stable within ±5% of the theoretical curvature. For example, cracks will cause local surface depressions / protrusions, and peeling will cause uneven surfaces. Real defects will cause the point cloud curvature to deviate from the theoretical value. Based on this, the lower limit of the curvature threshold for the defective candidate area is initially determined. Based on this, and combined with a historical health point cloud database of bridges and components of the same type and specifications, the initial threshold was preliminarily calibrated: At least 20 sets of point cloud curvature data for healthy components under the same working conditions were selected, and the mean and standard deviation of their curvature distribution were statistically analyzed. The upper limit of curvature in the healthy area was corrected to "theoretical curvature + 5% + 1 standard deviation". Simultaneously, point cloud samples of real defects of the same type were collected, and the mean curvature of the core defect area was extracted. The lower limit of curvature in the defect area was corrected to "theoretical curvature + 10%" or "theoretical curvature coefficient of variation ≥ 0.15". For different component types and defect types, the threshold was refined and set: For flat components, such as the web of a box girder and bridge deck, since the overall surface curvature is stable, a fixed numerical threshold was used, such as an upper limit of curvature in the healthy area ≤ 0.018m. -1 The lower limit of the curvature of the crack defect is ≥0.022m. -1 Finally, the threshold is finalized through pilot verification: For example, 3-5 pilot bridges with different service years and different types of defects are selected. The set curvature threshold is applied to point cloud curvature analysis to screen out candidate defect areas. This is then verified by combining on-site manual measurements with image recognition results. The true positive rate, false positive rate, and missed detection rate of the curvature threshold are statistically analyzed. If the false positive rate is higher than 3%, the lower limit of curvature for defect areas is appropriately increased; if the missed detection rate is higher than 2%, the lower limit of curvature for defect areas is appropriately decreased. This continues until the accuracy of the curvature threshold determination is ≥95%, the false positive rate is ≤3%, and the missed detection rate is ≤2%. This curvature threshold is then incorporated into the supporting parameter library of the bridge structure analysis standard as a unified basis for subsequent point cloud curvature analysis, false defect removal, and candidate defect area screening.

[0101] Therefore, by determining a curvature threshold, candidate disease regions can be screened. Based on the screened candidate regions, false diseases caused by stains, watermarks, shadows, etc., can be eliminated. Finally, the geometric features extracted from the image and the spatial coordinates obtained from the point cloud are integrated to form a standardized set of disease feature parameters that includes the spatial coordinates, geometric morphology parameters, and curvature features of the disease.

[0102] S23. Based on the set of defect feature parameters and combined with the bridge structure analysis standard, use the spatial inclusion relationship algorithm to determine the defect-related components, and obtain the defect component code based on the defect-related components;

[0103] In a preferred embodiment, the step of determining the associated components of a defect based on a set of defect feature parameters and in conjunction with bridge structure analytical standards, and obtaining the defect component code based on the associated components, includes the following steps:

[0104] S231. Based on the spatial coordinates of the defects in the defect feature parameter set, and combined with the bridge structure analysis standard, the point cloud coordinates of the defect area are transformed to the coordinate system of the bridge structure analysis model to obtain the transformed spatial coordinates of the defects.

[0105] It should be noted that the spatial coordinates of the disease in the disease characteristic parameter set include the local coordinates of the disease recorded based on temporary reference points during on-site acquisition, the three-dimensional coordinates of key points and the center of the disease boundary, and the global geodetic coordinates of the disease area synchronously acquired by GPS / RTK equipment. During the transformation, the global coordinates obtained by GPS / RTK are used as the control reference. The spatial similarity transformation algorithm is used to calculate and apply rotation, translation and scale parameters to transform the local coordinates of the disease into the global engineering coordinate system used by the bridge structure analytical model.

[0106] The process of calculating and applying rotation, translation, and scale parameters using a spatial similarity transformation algorithm includes the following steps: selecting three or more matching point pairs between the local coordinates of the disease and the corresponding global geodetic coordinates, using the least squares method to iteratively calculate key transformation parameters such as rotation angle, translation amount, and scale factor to eliminate positional deviations caused by temporary reference points; and then batch substituting the point cloud coordinates of all disease areas into the transformation model to achieve a unified transformation from local coordinates to the global engineering coordinate system.

[0107] S232. Using a pre-built spatial index structure and combined with the converted spatial coordinates of the disease, perform component queries to initially screen out the set of components whose spatial boundaries intersect with the disease area; and calculate the spatial relationship matching cost between the initially screened components and the disease area based on spatial relationship constraints, sort the components according to the matching cost, and select several components as a candidate component set.

[0108] In a preferred embodiment, the step of using a pre-built spatial index structure, combined with the converted spatial coordinates of the disease, to query components and initially filter out a set of components whose spatial boundaries intersect with the disease area; and calculating the spatial relationship matching cost between the initially filtered components and the disease area based on spatial relationship constraints, sorting the components according to the matching cost, and selecting several components as a candidate component set includes the following steps:

[0109] S2321. Based on the transformed spatial coordinates of the disease, a pre-built spatial index structure is used to query components, and all components intersecting with the disease area are determined by calculating the spatial bounding box, forming a preliminary set of selected components.

[0110] In a preferred embodiment, the step of querying components based on the transformed spatial coordinates of the disease using a pre-built spatial index structure, and determining all components intersecting with the disease area by calculating the spatial bounding box to form a preliminary set of selected components includes the following steps:

[0111] S23211. For each component in the analytical model of the bridge structure, calculate the minimum and maximum coordinate points of its spatial bounding box and store them in a pre-built spatial index structure.

[0112] It should be noted that when calculating the minimum and maximum coordinates of the bounding box, the minimum and maximum coordinate values ​​in the X, Y, and Z directions are extracted by traversing the three-dimensional coordinates of all vertices on the component's surface. This determines the scope of the component's bounding box and its minimum and maximum coordinates. The spatial index structure uses an R-tree. The minimum and maximum coordinates of each component's bounding box are treated as spatial data items, and these items are strongly associated with the component's unique identifier (i.e., its generated component code), forming a code pair. Then, the batch loading or sequential insertion method of the index structure is called to organize all component bounding boxes and code pairs into the tree-like hierarchical structure of the index. The R-tree algorithm automatically aggregates bounding boxes of components with similar spatial locations under the same intermediate node, forming a spatial hierarchy from the root node to the leaf nodes.

[0113] S23212. Convert the transformed set of spatial coordinates of the disease into a spatial bounding box of the disease area, and calculate the minimum and maximum coordinate points of the disease point cloud coordinates.

[0114] It should be noted that the process involves extracting the 3D coordinate data of all disease point clouds from the transformed disease spatial coordinate set, including the key points of the disease boundary, internal sampling points, and the coordinates of the region center. Then, by iterating through all disease point coordinates, the extreme values ​​in the X, Y, and Z directions are calculated: Xmin is the minimum X-coordinate of all disease points, Xmax is the maximum, and the Y and Z directions are calculated using the same logic. Using these six extreme values ​​as boundaries, a minimum spatial bounding box (MBR) that completely encloses all disease points is fitted and generated. The boundaries of this bounding box in the X, Y, and Z directions are then extended outwards by 5mm to form the final disease region bounding box used for querying.

[0115] S23213. Using a pre-built spatial index structure, the spatial bounding box intersection determination algorithm is used to query all components that intersect with the bounding box of the diseased area to obtain the query results.

[0116] Specifically, using a pre-built hierarchical R-tree spatial index, a two-level retrieval logic of upper-level coarse screening and lower-level fine screening is executed: First, using the generated bounding boxes of the diseased area as search conditions, the upper-level nodes of the index are queried. By comparing the coordinate range of the bounding boxes with those of various structural systems, structural systems that spatially intersect with the diseased area are quickly selected, directly excluding all components under non-intersecting systems, significantly narrowing the search scope. Second, for the selected target structural systems, the corresponding lower-level nodes of the index are retrieved, and the coordinates of the bounding boxes of all components under that system are extracted. Subsequently, a spatial bounding box intersection determination algorithm is used to verify the intersection relationship between the bounding boxes of components and the bounding boxes of the diseased area one by one: if the projections of two bounding boxes overlap in the X, Y, and Z axes, they are determined to be spatially intersecting, and the component is recorded as a candidate object. Finally, all intersecting components are integrated to form the preliminary query results.

[0117] S23214. Extract component codes from the query results and generate a preliminary set of filtered components containing the component codes and spatial bounding box coordinates of all intersecting components.

[0118] Specifically, the standard codes of all intersecting components are extracted based on the principle of unique component codes. Simultaneously, the spatial bounding box coordinates of each component are extracted, ensuring a complete correspondence between the identifier and spatial extent of each component in the set. This information is then integrated in a structured format, with fields including component code, structural system to which it belongs, minimum bounding box coordinates, and maximum bounding box coordinates. Components are then categorized and sorted according to their structural system to generate a preliminary filtered set of components.

[0119] S2322. For each component in the preliminary selected component set, extract its spatial relationship feature vector with the diseased area;

[0120] Specifically, the spatial relationship feature vector between each component and the diseased area is extracted. Four core spatial relationship features are extracted according to a unified rule, forming a 4-dimensional feature vector [f1, f2, f3, f4]. The calculation results of these four dimensions are then concatenated in a fixed order of f1, f2, f3, f4 to form the spatial relationship feature vector between each component and the diseased area. The calculation method for each dimension is as follows:

[0121] f1 represents the normalized center distance value. It is obtained by calculating the Euclidean distance D between the center of the component and the center of the defect, and normalizing D to the [0,1] interval based on the spatial range of the component's level in the bridge structure analysis standard. The closer the distance, the smaller the value of f1.

[0122] f2 represents the bounding box overlap rate. The overlap rate f2 (value 0-1) is obtained by calculating the intersection volume Vinter of the component bounding box and the defect bounding box and dividing it by the union volume Vunion of the two. The higher the overlap rate, the larger the value of f2.

[0123] f3 represents azimuth consistency. By taking the disease center as the origin, the azimuth angle θ1 of the component center is calculated, and the azimuth angle θ2 corresponding to the disease direction (based on the fitting of key points of the disease boundary) is calculated at the same time. The absolute value of the difference between the two azimuth angles is normalized to [0,1] to achieve azimuth consistency. The smaller the difference, the closer the f3 value is to 1.

[0124] f4 represents the size fit. It is calculated by the size ratio between the defect enclosure and the component enclosure, that is, the average of the ratios in the length, width and height directions, and normalized to [0,1] to obtain the size fit. The closer the ratio is to 1, the better the fit between the defect size and the local size of the component, and the larger the f4 value.

[0125] S2323. Based on the spatial relationship feature vector, use the preset matching cost function to calculate the spatial relationship matching cost between each component and the diseased area;

[0126] It should be noted that the preset matching cost function is a weighted summation model, and the function expression is:

[0127] Cost=w1×(1-f2)+w2×f1+w3×(1-f3)+w4×(1-f4);

[0128] In the formula, w1, w2, w3, and w4 are feature weight coefficients, which are set based on engineering experience and structural analysis standards: w1=0.4, w2=0.3, w3=0.15, w4=0.15, and w1+w2+w3+w4=1.

[0129] Specifically, by substituting the spatial relationship feature vector [f1,f2,f3,f4] of each component into the cost function, the matching cost (Cost) is calculated one by one. During the calculation, all values ​​are retained to four decimal places to ensure calculation accuracy. Finally, the matching cost corresponding to each component is obtained, with a value ranging from 0 to 1. The closer the Cost value is to 0, the closer the spatial association between the component and the defect.

[0130] S2324. Based on the calculated spatial relationship matching cost, sort the components in the preliminary screening component set in ascending order, and select several components with the smallest matching cost as the candidate component set.

[0131] Specifically, all components in the initial component set are initially screened and sorted in ascending order according to their corresponding matching cost (Cost), in other words, from minimum to maximum, forming a sorted component list. The list contains three core pieces of information: component code, matching cost, and spatial relationship feature vector. Then, the top 3-5 components with the lowest matching cost are selected first. The selected component information is then integrated to generate a candidate component set.

[0132] S233. For each candidate component, the point cloud surface matching algorithm is used to determine whether the point cloud coordinates of the diseased area are located on the surface of the component, and the overlap ratio between the point cloud of the diseased area and the surface of the component is generated based on the determination result.

[0133] In a preferred embodiment, for each candidate component, determining whether the point cloud coordinates of the defective area are located on the component surface using a point cloud surface matching algorithm, and generating the overlap ratio between the point cloud of the defective area and the component surface based on the determination result, includes the following steps:

[0134] S2331. For each candidate component, based on the transformed spatial coordinates of the disease and the surface data of the component, calculate the shortest spatial distance from each point in the point cloud of the diseased area to the surface of the component.

[0135] Specifically, for each candidate component, a KD-tree indexing algorithm is used to construct an index for the surface point cloud of the candidate component. The purpose is to transform the unordered surface point cloud into an ordered index structure, reducing the time complexity of nearest neighbor point retrieval. During the construction process, the segmentation dimension is adaptively selected based on the spatial distribution density of the component surface point cloud: if the point cloud is most dispersed in the X-axis direction, the X-axis is prioritized as the first segmentation dimension, and the median point under this dimension is selected as the segmentation node; subsequent levels alternately select the Y-axis and Z-axis as segmentation dimensions, repeating the process of selecting dimensions, finding median points, and segmenting the point cloud until the number of point clouds in each leaf node is less than a preset threshold (usually set to 5-10 points), thus completing the construction of the KD-tree index.

[0136] The process involves traversing each point in the defect point cloud subset obtained from the above screening, using its three-dimensional coordinates as the retrieval input, and performing nearest neighbor retrieval through a KD tree: starting from the root node of the KD tree, traversing downwards layer by layer along the segmentation dimension to quickly locate the leaf node to which the defect point belongs, and then filtering out the component surface points closest to the current defect point within the point cloud range of the leaf node and its adjacent nodes. Subsequently, the spatial distance between these two points is calculated using the three-dimensional Euclidean distance formula. This distance is the shortest spatial distance from the current defect point to the candidate component surface. Specifically, since the component surface point cloud has covered the entire surface of the component, and the nearest neighbor point is the point on the component surface that is closest to the defect point, the distance between the two points can be directly equivalent to the shortest distance from the defect point to the component surface.

[0137] S2332. Based on the preset distance tolerance threshold, points whose shortest spatial distance is less than or equal to the distance tolerance threshold are identified as defect points that overlap with the surface of the component.

[0138] It should be noted that the spatial accuracy requirements of the bridge structure analysis standard are: component coordinate deviation ≤ 3mm, origin positioning deviation ≤ 3mm; the system error of point cloud acquisition is ±2mm for laser scanner acquisition accuracy; the cumulative error of coordinate transformation is ≤ 3mm; and the natural roughness and minute morphological deviation of component surfaces are typically ≤ 1mm for concrete components and ≤ 0.5mm for steel components. Based on the above factors, the distance tolerance threshold can be uniformly set to 3mm. Then, each point in the defect point cloud subset is traversed, and its shortest spatial distance is compared with the preset distance tolerance threshold. If the distance is ≤ the threshold, the point is determined to be a defect point overlapping with the component surface and marked as a valid point; if the distance is > the threshold, it is determined to be a non-overlapping point and marked as an invalid point.

[0139] S2333. Count the number of disease points that are judged to be overlapping, and calculate the proportion of them to the total number of points in the disease area point cloud, as the overlap ratio between the candidate component and the disease area point cloud.

[0140] S234. Calculate the overlap ratio between the point cloud of the diseased area and the surface of each candidate component, and select the component with the highest overlap ratio that exceeds the preset threshold as the disease-related component.

[0141] Specifically, the preset threshold is set at 70%, based on the spatial accuracy requirements of bridge structure analysis standards, engineering inspection experience, and the surface morphology characteristics of components. Among these, the component with the highest overlap ratio is selected first. If its ratio exceeds the preset threshold, it is directly identified as a component associated with the defect. If the component with the highest ratio does not reach the preset threshold, a secondary verification is required based on the matching cost. If both conditions are met, it is still identified as an associated component. If not, it is marked as a defect location pending verification and requires manual intervention for confirmation. If multiple candidate components have the same overlap ratio and are all the highest values, such as defects at the junction of adjacent components where the overlap ratio of both components is 85%, then the shortest distance statistics and azimuth consistency are further compared to finally determine the associated component.

[0142] S235. Based on the disease-related components and the geometric features in the disease characteristic parameter set, determine the final disease component code.

[0143] Specifically, such as Figure 10As shown, from the identified disease-related components, their standard component codes are extracted. These codes serve as initial candidate values ​​for disease component codes, containing information about the component unit and surface to which the disease belongs. Then, the core geometric features from the disease feature parameter set are retrieved and their compatibility with the structural features of the related components is verified to ensure that the component corresponding to the code is logically consistent with the disease morphology: if the disease is a longitudinal crack, the related component should be a beam primarily subjected to bending, and the component type code in the code must match the beam component; if the code corresponds to a pier subjected to transverse forces, the related component screening results need to be reviewed; if the disease is localized spalling, its area must match the surface dimensions of the related component; if the disease extends across multiple component surfaces, the code of the component surface where the main body of the disease is located is used, and the codes of cross-component surface diseases and other related component surfaces are noted in the code remarks. After the geometric features are verified to be correct, the initial candidate code is locked as the final disease component code. If the verification finds a deviation, such as the component surface corresponding to the code not matching the disease morphology, the process is backtracked to S234 to re-select related components, or the matching details between the disease point cloud and the component surface are manually reviewed until the code is completely consistent with the actual attached component of the disease.

[0144] S24. Based on the disease metadata system, the support vector machine algorithm is used to classify the disease feature parameter set to obtain the disease type;

[0145] It should be noted that the Support Vector Machine (SVM) algorithm can be used to classify disease types, ensuring that the classification results are consistent with industry standards and preset criteria. Specifically, the classification target can be clearly defined by retrieving the three-level classification rules (major, intermediate, and minor categories) and corresponding standard disease type codes from the disease metadata system. Then, the disease feature parameter set is processed, standardizing geometric features, curvature features, and other parameters to the same dimension, forming a high-dimensional feature input vector to leverage the SVM algorithm's advantage in processing high-dimensional data. Next, the SVM model is trained based on a preset training sample set from the disease metadata system, optimizing the kernel function and penalty parameters to improve classification accuracy. The processed target disease feature parameter vector is then input into the trained SVM model, which infers and outputs the corresponding disease classification result. This result is then matched with the standard disease type codes from the disease metadata system, ultimately yielding a standardized and traceable disease type, providing a clear basis for subsequent risk level calculations.

[0146] S25. Based on the disease type and disease characteristic parameter set, calculate the disease risk level according to the quantitative grading threshold in the disease metadata system;

[0147] It should be noted that, based on the determined type of defect, the corresponding quantitative grading threshold is retrieved from the defect metadata system. This threshold has been differentiated by component material, stress type, and design load level, as shown in Table 1. Then, core parameters directly related to the grading threshold are extracted from the defect feature parameter set: maximum width for cracks and area and depth for spalling. Combined with the component attributes associated with the component code (material is C50 concrete, stress type is bending), the core parameters are weighted and corrected. The corrected core parameters are compared with the quantitative grading threshold one by one. According to the logic of parameter range matching and component attribute adaptation, the risk level corresponding to the defect is determined, including four levels: minor / moderate / relatively serious / serious. If the core parameter is in the threshold critical range, a secondary judgment is made based on the defect development trend characteristics, and finally, a defect risk level that fits the actual project is output.

[0148] Table 1 Disease Risk Level Determination Matrix

[0149] S26. The defects of the components, the types of defects, and the risk levels are encapsulated in a structured manner to generate spatial location data of highway bridge defects.

[0150] S3. Based on the spatial location data, construct a bridge defect model, and combine it with the bridge structural analytical model to construct a coordinate transformation system for spatial positioning of defects;

[0151] It should be noted that, based on the generated spatial location data, core information for each defect is extracted, including attribute information such as defect component code, type, and risk level, as well as spatial information such as local coordinates, global coordinates, and geometric morphological parameters. A modeling tool derived from the analytical model of the bridge structure is used, and parametric modeling methods are employed to construct the bridge defect model, such as... Figure 4 As shown or as Figure 14-15 As shown, each defect in the model is presented as an independent geometric entity and associated with a complete attribute dataset; then, by combining the coordinate system definition and coordinate transformation parameters of the bridge structure analytical model, a coordinate transformation system for spatial positioning of defects is constructed.

[0152] In constructing the coordinate transformation system for spatial positioning of defects, the global engineering coordinate system definition and coordinate transformation parameters explicitly defined in the bridge structure analytical model are retrieved. The core transformation logic is based on the local coordinates of defects, the local coordinate system of component surfaces, and the global engineering coordinate system. The local coordinates of defects based on the origin of the corresponding component surface are mapped to the local coordinate system of the component surface by matching the transformation parameters associated with the component code. Then, a secondary transformation is completed by combining the fixed coordinates of the component surface origin in the global engineering coordinate system. At the same time, the ICP iterative nearest point algorithm is introduced to compare the transformed defect point cloud with the surface point cloud of the corresponding component in the bridge structure analytical model. The transformation parameters are iteratively optimized to correct small deviations and ensure that the fit deviation between the spatial position of the defect and the component surface is ≤2mm. Finally, the complete transformation process, coordinate system parameters, verification rules, deviation thresholds, and automatic transformation interfaces adapted to subsequent defect data updates are standardized and encapsulated to form a coordinate transformation system for spatial positioning of defects.

[0153] S4. Based on the coordinate transformation system, the bridge defect model and the bridge structural analytical model are integrated to generate a bridge defect visualization model to present the location of defects.

[0154] Specifically, such as Figure 7-8 As shown, by using the component code of the defect as the core association key, matching the corresponding level of component units in the structural model, and calling the bidirectional transformation parameters and ICP iterative optimization algorithm in the coordinate transformation system, the independent geometric entities of the defects are accurately mapped to the corresponding spatial positions in the structural model, and all defect coordinates strictly follow the global engineering coordinate system standard of the structural model; then, a hierarchical linkage mechanism between the defect model and the structural model is established, so that each defect entity is automatically associated with a seven-level hierarchical structure of bridge, assessment unit, location, component, sub-component, and component surface according to its component code; at the same time, the complete attribute dataset of defect type, risk level, geometric morphology parameters, detection information, etc. are associated with the component attributes of the structural model to construct a structure- A two-way query link for bridge defects is established. Finally, differentiated visualization rendering rules are adopted, displaying defects according to type and risk level, and supplementing with defect annotation and scaling adaptation functions. This results in a visually appealing bridge defect model that intuitively presents the spatial distribution, type differences, and risk levels of defects within the bridge structure. Furthermore, the generated visual bridge defect model can be migrated to a digital twin system for use. Combined with vehicle monitoring on highway bridges, it enables real-time visualization and analysis of highway bridges. When the defect risk level escalates to relatively severe or severe, the system automatically generates precise maintenance work orders, which are directly pushed to the maintenance management platform and marked as pending in the digital twin, forming a closed-loop management system. Maintenance personnel can also access the digital twin system via mobile terminals to retrieve defect information corresponding to their current location.

[0155] S5. Based on the bridge defect visualization model and combined with historical defect data, track and compare the development status of defects, and evaluate the development trend of defects through the prediction model to generate bridge maintenance suggestions and repair plans.

[0156] Specifically, based on the current spatial distribution, type, risk level, and geometric parameters of bridge defects presented by the bridge defect visualization model, standardized inspection data of the same component code and type of defects from the historical defect database are retrieved. This includes location coordinates, dimensional parameters, and risk levels at different time points. Then, through the layer overlay and comparison functions of the bridge defect visualization model, the development status of defects is intuitively tracked, including changes in geometric shape, spatial location shifts, and risk level escalation. At the same time, the rate of defect development and changes in the scope of impact are quantitatively calculated. Finally, a time-series-based LSTM prediction model is used, incorporating environmental factors, component stress state, and structural health monitoring data, using the current core parameters of defects as input, to assess the next 1-5 years. The study analyzes the development trend of bridge defects to predict whether they will exceed safety thresholds, trigger structural risks, or escalate to a severe level. Finally, referencing the importance classification of components in the "Highway Bridge and Culvert Maintenance Specification" (JTG5120) and bridge structure analysis standards (e.g., based on stress type and functional priority), and combining the defect type, risk level, and development trend, differentiated bridge maintenance recommendations are generated. For example, quarterly monitoring plans are developed for minor-risk defects, repairs are scheduled within six months for medium-risk defects, and emergency treatment is initiated for severe-risk defects. The study also determines the selection of repair materials, key construction techniques, schedule, and quality acceptance standards. The plan simultaneously links the component codes and spatial location information corresponding to the defects, ensuring precise implementation of maintenance work and achieving scientific management of the entire defect lifecycle.

[0157] like Figure 9 As shown, according to a second embodiment of the present invention, a system for structured acquisition and tracking inspection of highway bridge defects is provided, the system comprising:

[0158] Module 1 is constructed to analyze components based on the pre-built bridge structure analytical model using BIM technology, and to analyze the pre-acquired highway bridge drawings, materials, and on-site measured data. Based on the component analysis results, a highway bridge structure analytical standard and defect metadata system is constructed.

[0159] The disease analysis module 2 is used to collect highway bridge disease data based on the highway bridge structure analysis standard and disease metadata system, and classify the disease data; based on the preset evaluation standard, the disease classification results are coded into components, and the risk level is divided by evaluating the degree of disease, and spatial location data including disease component code, disease type and risk level are obtained.

[0160] Coordinate transformation construction module 3 is used to construct a bridge defect model based on spatial location data, and to construct a coordinate transformation system for spatial positioning of defects in combination with the bridge structural analytical model;

[0161] The visualization model generation module 4 is used to integrate the bridge defect model with the bridge structure analytical model based on the coordinate transformation system to generate a bridge defect visualization model to present the location of the defect.

[0162] The disease prediction and assessment module 5 is used to track and compare the development status of bridge defects based on the bridge defect visualization model and historical defect data, and to assess the development trend of defects through the prediction model, and generate bridge maintenance suggestions and repair plans.

[0163] According to a third embodiment of the present invention, an electronic device is provided, the electronic device comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the steps in any of the above method embodiments.

[0164] According to a fourth embodiment of the present invention, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to perform the steps in any of the above method embodiments.

[0165] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0166] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0167] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for structured data collection and tracking inspection of highway bridge defects, characterized in that, The method includes the following steps: S1. Based on the pre-constructed bridge structure analytical model using BIM technology, perform component analysis on the pre-acquired highway bridge drawings, materials, and on-site measured data, and construct a highway bridge structure analytical standard and defect metadata system based on the component analysis results. S2. Based on the highway bridge structure analysis standard and the disease metadata system, collect highway bridge disease data and classify the disease data; based on the preset evaluation standard, encode the disease classification results into components, and classify the risk level by assessing the degree of disease, to obtain spatial location data containing disease component codes, disease types, and risk levels. S3. Based on the spatial location data, construct a bridge defect model, and combine it with the bridge structural analytical model to construct a coordinate transformation system for spatial positioning of defects; S4. Based on the coordinate transformation system, the bridge defect model and the bridge structural analytical model are integrated to generate a bridge defect visualization model to present the location of defects. S5. Based on the bridge defect visualization model and combined with historical defect data, track and compare the development status of defects, and evaluate the development trend of defects through the prediction model to generate bridge maintenance suggestions and repair plans.

2. The method for structured data collection and tracking inspection of highway bridge defects according to claim 1, characterized in that, The process of constructing a bridge structural analytical model using BIM technology, analyzing pre-acquired highway bridge drawings, materials, and on-site measured data to identify components, and building a highway bridge structural analytical standard and defect metadata system based on the component analysis results includes the following steps: S11. Obtain highway bridge drawings and materials and on-site measurement data, and based on the highway bridge drawings and materials, use BIM software to construct an analytical model of the bridge structure; S12. Perform component analysis on the analytical model of the highway bridge structure, and determine the origin coordinate system of the component surface by extracting component features, forming a construction analysis result that includes component coding rules and component hierarchical relationships; S13. Based on the component analysis results, configure the bridge structure analysis standard, which includes component coding rules, origin coordinate system definition, component hierarchical relationship and spatial position requirements; S14. Based on the predefined disease classification and severity assessment rules, generate a disease metadata system that includes standard disease type codes, disease attribute field definitions, quantitative grading thresholds, and spatial data record specifications associated with component codes and the origin coordinate system.

3. The method for structured data collection and tracking inspection of highway bridge defects according to claim 1, characterized in that, The process involves collecting highway bridge defect data based on the highway bridge structure analysis standard and defect metadata system, classifying the defect data, encoding the defect classification results into components based on preset evaluation standards, and classifying risk levels by assessing the degree of defect, resulting in spatial location data containing defect component codes, defect types, and risk levels. This includes the following steps: S21. Use data acquisition equipment to acquire point cloud data and image data of the surface of highway bridges; S22. Preprocess the image data using an image recognition algorithm to extract disease feature parameters; and verify the disease area through point cloud curvature analysis to obtain a set of disease feature parameters containing disease spatial coordinates and disease geometric features. S23. Based on the set of defect feature parameters and combined with the bridge structure analysis standard, use the spatial inclusion relationship algorithm to determine the defect-related components, and obtain the defect component code based on the defect-related components; S24. Based on the disease metadata system, the support vector machine algorithm is used to classify the disease feature parameter set to obtain the disease type; S25. Based on the disease type and disease characteristic parameter set, calculate the disease risk level according to the quantitative grading threshold in the disease metadata system; S26. The defects of the components, the types of defects, and the risk levels are encapsulated in a structured manner to generate spatial location data of highway bridge defects.

4. The method for structured data collection and tracking inspection of highway bridge defects according to claim 3, characterized in that, The process of determining the associated components of a defect based on a set of defect feature parameters and in conjunction with bridge structure analysis standards, using a spatial inclusion relationship algorithm, and obtaining the defect component codes based on these associated components includes the following steps: S231. Based on the spatial coordinates of the defects in the defect feature parameter set, and combined with the bridge structure analysis standard, the point cloud coordinates of the defect area are transformed to the coordinate system of the bridge structure analysis model to obtain the transformed spatial coordinates of the defects. S232. Using a pre-built spatial index structure and combined with the converted spatial coordinates of the disease, perform component queries to initially screen out the set of components whose spatial boundaries intersect with the disease area; and calculate the spatial relationship matching cost between the initially screened components and the disease area based on spatial relationship constraints, sort the components according to the matching cost, and select several components as a candidate component set. S233. For each candidate component, the point cloud surface matching algorithm is used to determine whether the point cloud coordinates of the diseased area are located on the surface of the component, and the overlap ratio between the point cloud of the diseased area and the surface of the component is generated based on the determination result. S234. Calculate the overlap ratio between the point cloud of the diseased area and the surface of each candidate component, and select the component with the highest overlap ratio that exceeds the preset threshold as the disease-related component. S235. Based on the disease-related components and the geometric features in the disease characteristic parameter set, determine the final disease component code.

5. The method for structured data collection and tracking inspection of highway bridge defects according to claim 4, characterized in that, The pre-built spatial index structure is used in conjunction with the converted spatial coordinates of the disease to query components and initially filter out the set of components whose spatial boundaries intersect with the disease area; Based on spatial relationship constraints, the spatial relationship matching cost between the initially screened components and the diseased area is calculated. The components are then sorted according to the matching cost, and several components are selected as a candidate component set. This includes the following steps: S2321. Based on the transformed spatial coordinates of the disease, a pre-built spatial index structure is used to query components, and all components intersecting with the disease area are determined by calculating the spatial bounding box, forming a preliminary set of selected components. S2322. For each component in the preliminary selected component set, extract its spatial relationship feature vector with the diseased area; S2323. Based on the spatial relationship feature vector, use the preset matching cost function to calculate the spatial relationship matching cost between each component and the diseased area; S2324. Based on the calculated spatial relationship matching cost, sort the components in the preliminary screening component set in ascending order, and select several components with the smallest matching cost as the candidate component set.

6. The method for structured data collection and tracking inspection of highway bridge defects according to claim 5, characterized in that, The process of querying components based on the transformed spatial coordinates of the disease area, using a pre-built spatial index structure, and determining all components intersecting with the disease area by calculating the spatial bounding box to form a preliminary set of selected components includes the following steps: S23211. For each component in the analytical model of the bridge structure, calculate the minimum and maximum coordinate points of its spatial bounding box and store them in a pre-built spatial index structure. S23212. Convert the transformed set of disease spatial coordinates into a spatial bounding box of the disease area, and calculate the minimum and maximum coordinate points of the disease point cloud coordinates. S23213. Using a pre-built spatial index structure, and through a spatial bounding box intersection determination algorithm, query all components that intersect with the bounding box of the diseased area to obtain the query results; S23214. Extract component codes from the query results and generate a preliminary set of filtered components containing the component codes and spatial bounding box coordinates of all intersecting components.

7. The method for structured data collection and tracking inspection of highway bridge defects according to claim 5, characterized in that, For each candidate component, the process of determining whether the point cloud coordinates of the defective area lie on the component surface using a point cloud surface matching algorithm, and generating the overlap ratio between the point cloud of the defective area and the component surface based on the determination result, includes the following steps: S2331. For each candidate component, based on the transformed spatial coordinates of the disease and the surface data of the component, calculate the shortest spatial distance from each point in the point cloud of the diseased area to the surface of the component. S2332. Based on the preset distance tolerance threshold, points whose shortest spatial distance is less than or equal to the distance tolerance threshold are identified as defect points that overlap with the surface of the component. S2333. Count the number of disease points that are judged to be overlapping, and calculate the proportion of them to the total number of points in the disease area point cloud, as the overlap ratio between the candidate component and the disease area point cloud.

8. A system for structured acquisition and tracking inspection of highway bridge defect data, used to implement the method for structured acquisition and tracking inspection of highway bridge defect data as described in any one of claims 1-7, characterized in that, The system includes: The analysis module is used to analyze the components of the pre-built bridge structure analysis model using BIM technology, and to construct a highway bridge structure analysis standard and a defect metadata system based on the pre-acquired highway bridge drawings, materials and on-site measured data. The disease analysis module is used to collect highway bridge disease data based on the highway bridge structure analysis standard and disease metadata system, and classify the disease data; based on the preset evaluation standard, the disease classification results are coded into components, and the risk level is divided by assessing the degree of disease, resulting in spatial location data containing disease component codes, disease types, and risk levels. The coordinate transformation construction module is used to build a bridge defect model based on spatial location data, and combine it with the bridge structural analytical model to build a coordinate transformation system for spatial positioning of defects; The visualization model generation module is used to fuse the bridge defect model with the bridge structure analytical model based on the coordinate transformation system, and generate a visualization model of bridge defects to show the location of defects. The bridge damage prediction and assessment module is used to track and compare the development status of bridge damage based on a bridge damage visualization model and historical damage data. It also uses the prediction model to assess the development trend of damage and generate bridge maintenance suggestions and repair plans.

9. An electronic device, characterized in that, The electronic device includes: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the steps of the method according to any one of claims 1 to 7 are implemented when the computer program controls the device containing the computer-readable storage medium to execute during runtime.

Images