A method and system for constructing a super-large anchor digital twin

Abstract

The application discloses a kind of super large anchorage digital twin construction method and system, the method includes the following steps: arrangement 16 hole inclinometer tube forms three-dimensional monitoring array, real-time data acquisition is carried out using inclinometer, accurate grid point data is generated using IDW interpolation algorithm;Optimization processing is carried out to digital model, including structure optimization, efficient compression algorithm application and instantiation rendering;Realize three-dimensional visualization based on WebGL technology, including model translucency, grid data accurate alignment, custom shader realizes heat map rendering etc.;Build a comprehensive digital twin system platform, realize the whole life cycle management and real-time monitoring of super large anchorage structure.The application collects real-time monitoring mode by arranging sensor, significantly improves the accuracy and comprehensiveness of data acquisition and realizes the real-time monitoring of anchorage global deformation, realizes the high-precision, real-time, interactive digital twin visualization of super large anchorage structure, provides strong support for engineering monitoring and decision-making.

Description

Technical Field

[0001] This invention relates to the fields of civil engineering monitoring and bridge structure digital twin technology, and in particular to a method and system for constructing a digital twin of an ultra-large anchorage. Background Technology

[0002] With the rapid development of modern transportation infrastructure, large-scale bridge projects are becoming increasingly complex and intelligent. As a key load-bearing structure of a bridge, the safety and stability of anchorages directly affect the service life and operational safety of the entire bridge. However, traditional monitoring methods have significant limitations: high-precision monitoring methods are often inefficient and costly, while high-efficiency methods may lack accuracy and have large deviations. More importantly, these methods mostly acquire data from discrete points, making it difficult to comprehensively reflect the overall performance of the anchorages. In addition, many traditional methods also face problems such as insufficient real-time performance, significant influence from environmental factors, and difficulty in achieving long-term continuous monitoring, failing to meet the needs of modern bridge engineering for accurate, real-time, and comprehensive monitoring.

[0003] Therefore, there is an urgent need to develop an innovative monitoring system to meet the efficient management requirements of modern bridge engineering. This system, based on digital twin technology, should possess characteristics such as real-time global performance monitoring, high-precision visualization, efficient data processing, intelligent early warning mechanisms, remote monitoring capabilities, and system integration. It can capture the overall deformation and stress distribution of the anchorage structure in real time, providing an intuitive and accurate three-dimensional visualization interface, rapidly processing and analyzing large amounts of monitoring data, and achieving intelligent early warning of potential risks. Simultaneously, it supports remote access and control, facilitating remote diagnosis and decision-making by experts. This system is not only an inevitable trend in technological development but will also provide strong technical support for the full life-cycle management of bridges, promoting the development of bridge engineering towards a more intelligent and safer direction. Summary of the Invention

[0004] To address the aforementioned shortcomings or improvement needs of existing technologies, this invention aims to resolve the limitations of traditional monitoring methods: high-precision monitoring methods are often inefficient and costly, while high-efficiency methods may lack accuracy and contain significant biases. More importantly, these methods primarily acquire data from discrete points, making it difficult to comprehensively reflect the overall performance of the anchorage. Furthermore, many traditional methods suffer from insufficient real-time performance, significant influence from environmental factors, and difficulty in achieving long-term continuous monitoring, thus failing to meet the demands of modern bridge engineering for accurate, real-time, and comprehensive monitoring.

[0005] This invention provides a method and system for constructing a digital twin of an ultra-large anchorage, comprising the following steps:

[0006] Step S1. Set up a three-dimensional monitoring array of inclinometer tubes that are evenly distributed in the horizontal and vertical directions on the anchorage diaphragm wall structure, and transmit the collected displacement data to the field data processing center in real time through an information transmission device;

[0007] Step S2. Using the 2D surface unfolding interpolation method, the three-dimensional structural features of the anchorage diaphragm wall are converted into two-dimensional coordinates. In the unfolded two-dimensional coordinate system, the two-dimensional plane inverse distance interpolation method is used to complete the grid data interpolation expansion, and the grid point displacement values ​​and their corresponding coordinate data are stored in the database in real time in matrix form.

[0008] Step S3. Based on the structural data of the anchorage diaphragm wall recorded during the design and construction process, create a BIM model, optimize the model structure and compress the model file size, and then reduce the computer memory usage and rendering calls during model loading and rendering by repeating component instantiation to obtain a lightweight model;

[0009] Step S4. Initialize the 3D scene, configure the orthogonal camera to adapt to the large anchorage structure; set ambient light and parallel light source to enhance the sense of three-dimensionality; import the lightweight model after step S3 and configure the WEB 3D scene, align the grid data stored in the database according to step S2 with the BIM model coordinate system, and then process the grid data through the shader to complete the heat map rendering and visualization.

[0010] Step S5. Integrate multi-source heterogeneous data, including real-time monitoring data from sensor networks, BIM model data, and historical construction data, and display the monitoring data in a fusion manner to provide a multi-dimensional visualization interface.

[0011] Furthermore, the three-dimensional monitoring array in step S1 is: a three-dimensional monitoring array of inclinometer tubes with no less than 16 holes uniformly distributed along the horizontal and vertical directions, the depth of the inclinometer tubes being consistent with the diaphragm wall structure, the guide groove of the inclinometer tubes being perpendicular to the edge of the pit and pointing towards the center of the pit, and the sensor acquisition frequency in the inclinometer tubes being 1 time / hour.

[0012] Furthermore, the information transmission device in step S1 includes a distributed optical fiber sensor demodulator, an industrial-grade switch, and an optical fiber network.

[0013] Furthermore, the 2D surface unfolding interpolation method in step S2 includes:

[0014] (1) Coordinate system establishment: Set the central axis of the diaphragm wall cylinder as the z-axis of the Cartesian coordinate system, with the center of the cylinder ground as the origin;

[0015] (2) Coordinate transformation: Replace the Cartesian sensor coordinates and values ​​(x,y,z) with cylindrical coordinates (r,θ,z); where r is the distance from the sensor to the central axis, θ is the angle of the sensor on the horizontal plane, θ∈[0,2π);

[0016] (3) Boundary condition continuity processing: To ensure continuity at θ = 0 and θ = 2π, that is, the continuity of the left and right boundaries after the cylinder is unfolded, the data at θ = 0 is copied at θ = 2π.

[0017] (4) Coordinate normalization: The angle θ is normalized according to the actual arc length of the cylindrical surface, that is, x_θ=θ×r, the x-axis represents the distance along the circumference, and the unit is the same as the z-axis (height);

[0018] (5) The calculation formula is as follows: (Using the two-dimensional inverse plane distance interpolation method)

[0019]

[0020]

[0021] In the formula It is the estimated value of position x0, Z(x i ) is a known position x i The observed values, and the predicted location of the unknown point (x0, y0). It is all known points (x) i y i The weighted sum of the values ​​Z is the sum of the values ​​of the unknown points. The weight of each known point is determined by the reciprocal of its distance di from the unknown point to the i-th known point, raised to the power of α.

[0022] Furthermore, the specific methods for optimizing the model structure in step S3 include: deleting unnecessary internal components, removing hidden surfaces and overlapping geometry, merging coplanar small patches, simplifying complex surfaces, reducing visually insignificant areas of surface, organizing the model tree structure, and deleting redundant groups and subgroups to flatten the model structure.

[0023] Furthermore, the specific method for compressing the model file size in step S3 includes: using the Draco compression algorithm to quantize vertex coordinates, converting floating-point numbers to integers to reduce data precision while maintaining visual quality; using predictive coding technology to compress vertex data, utilizing the spatial relationships between adjacent vertices to reduce data redundancy; encoding the topology to reduce redundant information; and setting a compression level to complete model size compression while maintaining decoding speed.

[0024] Furthermore, the specific method for the shader to process data in step S4 includes:

[0025] (1) Data normalization processing: Calculate the spatial coordinates and the given normalized displacement value v. Assume the minimum displacement value v_min and the maximum displacement value v_max. The normalization formula is:

[0026] t = (v - v_min) / (v_max - v_min)

[0027] (2) Using an additive mixing approach, for each pixel in the overlapping region, a new color value C_result is assigned:

[0028] C_result=min(C1+C2+...+Cn,1.0)

[0029] Where C1, C2, ..., Cn are the color values ​​of the n overlapping points at that pixel position;

[0030] Furthermore, step S5 provides multi-dimensional visualization interface functions, including providing a multi-dimensional visualization interface, including 3D model view, 2D cross-sectional view and time series data chart.

[0031] As another aspect of the present invention, a digital twin construction system for ultra-large anchorages is also provided, including a real-time data acquisition unit for large anchorage structures, a grid data generation unit, a lightweight BIM processing unit, a real-time rendering unit for three-dimensional point clouds, and a digital twin platform unit.

[0032] The large anchor structure real-time data acquisition unit is used to acquire raw sensor displacement data in real time;

[0033] The grid data generation unit is used to expand the collected raw data and enrich the grid point data set;

[0034] The BIM lightweight processing unit is used to construct a digital model of a large anchorage structure and optimize the model to reduce information redundancy, making the model lightweight and efficient in actual operation.

[0035] The three-dimensional point cloud real-time rendering unit is used to color the digital information model of the anchorage structure.

[0036] The digital twin platform unit is used to integrate all multi-source heterogeneous data and produce visualization graphics to realize full life cycle management and real-time monitoring of ultra-large anchorage structures.

[0037] As another aspect of the invention, a computer-readable storage medium is also disclosed, on which a computer program is stored, which is executed by a processor as a method for constructing a digital twin of an ultra-large anchorage for monitoring large-scale bridge engineering projects.

[0038] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0039] 1. The method for constructing a digital twin of an ultra-large anchorage of the present invention significantly improves the accuracy and comprehensiveness of data acquisition and realizes real-time monitoring of the global deformation of the anchorage by arranging sensors for real-time monitoring and deploying a high-speed data acquisition and transmission network.

[0040] 2. The method for constructing a digital twin of an ultra-large anchorage in this invention significantly reduces the amount of model data and improves rendering efficiency through BIM lightweight processing, enabling complex models to run smoothly on ordinary devices.

[0041] 3. The method for constructing a digital twin of a super-large anchorage in this invention uses WebGL technology to achieve real-time rendering of large-scale three-dimensional point clouds, which greatly improves the visualization effect and interactive experience; the web-based visualization system facilitates multi-terminal access and remote monitoring, improving monitoring efficiency and decision-making speed; the constructed digital twin platform provides strong support for the safe operation of the bridge and helps to promptly detect and handle potential risks. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the overall process of a preferred embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of a preferred embodiment of the three-dimensional column two-dimensional unfolding method of the present invention;

[0044] Figure 3 This is a schematic diagram of a BIM model according to a preferred embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of the completed 3D rendering of a preferred embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of a digital twin platform system according to a preferred embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0048] Please refer to Figure 1 The present invention provides a method for constructing a digital twin of an ultra-large anchorage, comprising:

[0049] 1) Data Acquisition and Grid Data Generation

[0050] Based on the geometric characteristics and stress features of the anchored diaphragm wall structure, a 16-well inclinometer array is uniformly distributed along the horizontal and vertical directions to form a three-dimensional monitoring system. The depth of the inclinometer tubes is consistent with the diaphragm wall structure, and the guide grooves are perpendicular to the edge of the pit and point towards the center of the pit. The sensor acquisition frequency is 1 time / hour. A distributed fiber optic sensor demodulator is used to achieve synchronous high-speed acquisition from all sensors. The acquired data is transmitted in real time to the field data processing center via an industrial-grade switch and fiber optic network.

[0051] Furthermore, in some other preferred embodiments, the types and functions of sensors are further expanded. In addition to inclinometer sensors, more types of sensors, such as stress sensors, strain sensors, and temperature sensors, are introduced to obtain more comprehensive anchorage structure status information. Simultaneously, the sensor placement strategy is optimized. The optimal sensor distribution scheme is determined through methods such as numerical simulation or physical experiments to improve the accuracy and reliability of data acquisition. Moreover, more advanced data transmission technologies, such as 5G communication technology or low-power wide-area network technology, are explored to further improve the speed and stability of data transmission, ensuring zero-delay transmission of real-time monitoring data.

[0052] Please refer to Figure 2 In numerical processing methods, 3D spatial interpolation is performed directly in a three-dimensional Cartesian coordinate system (x, y, z), preserving the original spatial information of the diaphragm wall. 2D surface unfolding interpolation "unfolds" the cylindrical diaphragm wall into a two-dimensional plane. The circumferential coordinates are mapped to the x-axis, representing the position along the circumference, and the height direction is mapped to the y-axis, preserving the original height information. Surface interpolation is performed on this (x, y) plane. The consideration for using the 2D unfolding method is that since sensors are deployed on the continuous curved surface of the diaphragm wall, 2D unfolding can better capture the spatial relationships and constraints between sensors on the surface, which may be ignored in pure 3D interpolation. The 2D plane unfolding method is as follows:

[0053] (1) Coordinate system establishment: Set the central axis of the diaphragm wall cylinder as the z-axis of the Cartesian coordinate system, with the center of the cylinder ground as the origin;

[0054] (2) Coordinate transformation: Replace the Cartesian sensor coordinates and values ​​(x, y, z) with cylindrical coordinates (r, θ, z). Where r is the distance from the sensor to the central axis, which should be a constant in this example and can be omitted, and θ is the angle of the sensor on the horizontal plane, θ∈[0, 2π).

[0055] (3) Boundary condition continuity processing: To ensure continuity at θ = 0 and θ = 2π (i.e., the continuity of the left and right boundaries after the cylinder is unfolded), the data at θ = 0 is copied at θ = 2π;

[0056] (4) Coordinate normalization: The angle θ is normalized according to the actual arc length of the cylindrical surface, that is, x_θ=θ×r, where the x-axis represents the distance along the circumference and the unit is the same as the z-axis (height).

[0057] The calculation formula is as follows: (Using the two-dimensional inverse distance interpolation method)

[0058]

[0059]

[0060] In the formula It is the estimated value of position x0, Z(x i ) is a known position x i The observed values, and the predicted location of the unknown point (x₀, y₀). It is all known points (x) i y i The weighted sum of the values ​​Z is the sum of the values ​​of the unknown points. The weight of each known point is determined by the reciprocal of its distance di from the unknown point to the i-th known point, raised to the power of α.

[0061] In this case, the circular curved diaphragm wall is divided vertically into 25 parts and circumferentially into 180 parts, with a grid size of approximately 2.5m × 2.5m, totaling 24 * 179 = 4296 grid points.

[0062] The grid point displacement values ​​and their corresponding coordinate data are stored in real time in a matrix format in the MongoDB database.

[0063] (2) Lightweight BIM processing

[0064] Please refer to Figure 3 A high-precision BIM model of the anchorage was created using Autodesk Revit software. The model underwent structural optimization, removing unnecessary internal components such as temporary supports and construction auxiliary structures. Hidden faces and overlapping geometry were removed to reduce redundant faces. Coplanar small faces were merged to reduce the total number of faces. Complex surfaces were simplified by approximating them with fewer polygons while maintaining acceptable geometric accuracy. For areas with minimal visual impact, surface accuracy could be appropriately reduced. The model hierarchy was optimized, and the model tree structure was rationally organized for easier subsequent selection and manipulation. Redundant groups and subgroups were removed to flatten the model structure.

[0065] The open-source Draco compression algorithm is used to quantize vertex coordinates, converting floating-point numbers to integers to reduce data precision while maintaining visual quality. Predictive coding is used to compress vertex data, leveraging the spatial relationships between adjacent vertices to reduce data redundancy. The topology is encoded to reduce redundant information. An appropriate compression level is set to strike a balance between file size and decoding speed. The compressed model file size can be reduced by approximately 70%-80% while maintaining acceptable visual quality.

[0066] Identify repetitive components in the model, such as prestressed anchors and reinforcing bars. Convert these repetitive components into instances, retaining a complete set of geometric data as a prototype. For other identical components, only position, rotation, and scaling information is stored. Implement instance rendering technology, rendering multiple instances in a single draw call during rendering, significantly reducing memory usage and the number of rendering calls.

[0067] The optimized BIM model is exported to glTF (GL Transmission Format), a format suitable for Web3D rendering. It supports compressed geometric data, can be efficiently streamed, and is widely supported by Web3D libraries such as Three.js.

[0068] In the lightweight processing of BIM models, more efficient model optimization algorithms and compression methods can be explored. For example, combining artificial intelligence technologies, such as model compression algorithms in deep learning, can automatically optimize and compress BIM models while maintaining their accuracy and integrity. For grid data generation and storage, more accurate interpolation methods or numerical simulation methods can be considered to improve the accuracy and resolution of grid data. Furthermore, research can be conducted on how to better integrate model data from different sources and with varying levels of precision to construct more accurate digital twins.

[0069] (3) Real-time rendering of 3D point clouds

[0070] 1) Scene initialization and BIM model import

[0071] The 3D scene is initialized using the Three.js library based on WebGL 2.0, and an orthographic camera is configured to adapt to the large anchored structure. Ambient light and parallel light sources are set to enhance the sense of depth. An optimized BIM model is imported, and a MeshPhongMaterial material with an opacity of 0.5 is applied to achieve a semi-transparent effect. The model is loaded using THREE.GLTFLoader to ensure efficient rendering.

[0072] 2) Grid data processing and coordinate alignment

[0073] The monitored grid data is converted to the BufferGeometry format of Three.js, and the position and displacement information are stored using a Float32Array. A precise coordinate transformation algorithm is implemented to align the grid data with the BIM model coordinate system, with the error controlled within 1cm. Matrix4 from Three.js is used for coordinate transformation to ensure high-precision alignment.

[0074] 3) Heatmap rendering and visualization

[0075] Please refer to Figure 4 Create a ShaderMaterial to implement custom shading. The vertex shader processes position and displacement data, calculates the spatial coordinates and a given normalized displacement value v, assuming a minimum value v_min and a maximum value v_max, the normalization formula is:

[0076] t = (v - v_min) / (v_max - v_min)

[0077] In point cloud heatmap rendering, when multiple points with blurred edges overlap, simple transparency overlay may lead to color saturation or distortion, and the boundaries of the overlapping areas may be blurred, making it difficult to accurately express the density of points.

[0078] To solve this problem, an additive blending approach is applied, where for each pixel in the overlapping region, the new color value is:

[0079] C_result=min(C1+C2+...+Cn,1.0)

[0080] Where C1, C2, ..., Cn are the color values ​​of the n overlapping points at that pixel position.

[0081] The specific implementation method is as follows: map the displacement value to the color space and set an appropriate transparency of 0.3. The rendering process is as follows: a. enable additive blending mode, b. render each point in sequence according to its spatial position, c. for each pixel, the GPU automatically performs additive blending, d. apply tone mapping or gamma correction to enhance the visual effect.

[0082] (4) Digital Twin System Platform

[0083] Please refer to Figure 5 Embodiments of this invention provide a digital twin platform that integrates multi-source heterogeneous data, including real-time monitoring data from sensor networks, BIM model data, and historical construction data. It employs distributed data storage technology and uses NoSQL databases such as MongoDB to achieve efficient large-scale data management.

[0084] Based on WebGL technology, this system integrates BIM models and monitoring data for seamless display. It provides a multi-dimensional visualization interface, including 3D model views, 2D cross-sectional views, and time-series data charts. User-defined views and interactive operations are supported, such as point querying, historical data playback, and early warning information display. The server periodically processes sensor data to generate updated grid data. Clients request updated data in JSON format via FileLoader, dynamically refreshing the BufferGeometry property. InstancedBufferGeometry is used for point instantiation rendering, and Web Workers handle computationally intensive tasks. Rendering resolution and refresh strategies are optimized to improve performance on different devices.

[0085] The digital twin system platform can be further expanded in functionality, such as by adding a structural health assessment module to perform real-time assessment and prediction of the anchorage structure's health status based on collected data and model analysis results. Simultaneously, an intelligent decision support system can be developed to automatically generate maintenance recommendations and decision-making schemes based on the structural health assessment results. Furthermore, strengthening integration and interaction with other bridge management systems will enable data sharing and collaborative management, improving the efficiency and level of bridge lifecycle management. Virtual reality (VR) and augmented reality (AR) technologies can also be utilized to provide users with a more immersive interactive experience, facilitating more intuitive remote diagnosis and decision-making by experts.

[0086] Example 2

[0087] This invention also proposes a system for constructing a digital twin of an ultra-large anchorage, comprising:

[0088] Real-time Data Acquisition Unit for Large Anchorage Structure: This unit is the data source for the entire system. It acquires data through a carefully designed three-dimensional monitoring array of inclinometer tubes on the anchorage diaphragm wall structure. This array is evenly distributed horizontally and vertically, with at least 16 inclinometer tubes. The depth of the inclinometer tubes is consistent with the diaphragm wall structure, and their guide grooves are perpendicular to the edge of the pit and point towards the center of the pit. The sensor acquisition frequency is once per hour, ensuring the scientific validity and effectiveness of the acquired data. Simultaneously, it is equipped with information transmission devices including a distributed fiber optic sensor demodulator, an industrial-grade switch, and a fiber optic network, enabling accurate and real-time transmission of the acquired raw sensor displacement data to the on-site data processing center, providing fundamental data support for subsequent data processing and analysis.

[0089] Grid Data Generation Unit: This unit is primarily responsible for processing and expanding the collected raw data. It employs a 2D surface unfolding interpolation method, first converting the Cartesian sensor coordinates and values ​​(x, y, z) into a cylindrical coordinate system (r, θ, z). After handling boundary condition continuity (copying data from θ = 0 at θ = 2π) and coordinate normalization, it uses a two-dimensional inverse distance interpolation method in the unfolded two-dimensional coordinate system. In this way, grid data interpolation and expansion are completed, enriching the grid point data set. The grid point displacement values ​​and their corresponding coordinate data are stored in real-time in matrix form in the database, laying the data foundation for subsequent integration with the BIM model and visualization.

[0090] The BIM Lightweight Processing Unit: This unit is responsible for building and optimizing the digital model of large anchorage structures within the system. After creating the BIM model based on the structural data of the anchorage diaphragm walls recorded during design and construction, a series of optimization measures are implemented to reduce model information redundancy. These include deleting unnecessary internal components, removing hidden faces and overlapping geometry, merging coplanar small patches, simplifying complex surfaces, reducing the visual impact of surfaces in areas with minimal impact, organizing the model tree structure, and deleting redundant groups and subgroups to flatten the model structure. Furthermore, the Draco compression algorithm is used to compress the model file size, quantize vertex coordinates, convert floating-point numbers to integers to reduce data precision while maintaining visual quality; predictive coding technology is used to compress vertex data, utilizing the spatial relationships between adjacent vertices to reduce data redundancy; and the topology is encoded to reduce redundant information. Appropriate compression levels are also set to make the model lightweight and efficient in actual operation, facilitating smooth operation and processing on different devices.

[0091] Real-time 3D Point Cloud Rendering Unit: This unit focuses on coloring the constructed digital information model of the anchorage structure. After initializing the 3D scene, an orthographic camera is configured to adapt to the large anchorage structure, while ambient light and parallel light sources are set to enhance the sense of depth. The lightweight BIM model is imported and a web-based 3D scene is configured, aligning the grid data stored in the database based on the grid data generation unit with the BIM model coordinate system. Data is processed through specific shaders, including data normalization (calculating the normalized values ​​of spatial coordinates and given displacements using the normalization formula: t = (v - v_min) / (v_max - v_min)) and an additive blending approach (assigning a new color value C_result = min(C1 + C2 + ... + Cn, 1.0) to complete heatmap rendering and visualization, improving the model's visualization effect and interactive experience.

[0092] Digital Twin Platform Module: The digital twin platform module is the core integration component of the entire system. It integrates heterogeneous data from multiple sources, including real-time monitoring data from sensor networks, BIM model data, and historical construction data, and performs comprehensive processing on this data. By fusing and displaying this data, it provides a multi-dimensional visualization interface, including 3D model views, 2D cross-sectional views, and time-series data charts, enabling full lifecycle management and real-time monitoring of ultra-large anchorage structures. This allows relevant personnel to comprehensively understand the status of the anchorages from multiple perspectives, facilitating timely detection of potential risks and decision-making, and providing strong support for the safe operation of large-scale bridge projects.

[0093] The ultra-large anchorage digital twin construction system of this invention can assist in design optimization and construction guidance in new bridge projects in the field of bridge engineering, and enable health monitoring and performance evaluation in the maintenance of existing bridges; in the field of civil engineering, it can be used for monitoring large building foundations and underground engineering; in the field of water conservancy engineering, it is suitable for monitoring water conservancy facilities such as dams, sluices and pumping stations; in the field of geological engineering, it can be used for geological disaster monitoring and early warning, as well as monitoring underground resource development; and in the fields of scientific research and teaching, it can be used as a scientific research experimental means and teaching and training tool. It has a wide range of applications and is of great significance for ensuring the safety of various engineering structures and promoting the development of related disciplines.

[0094] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of any of the above-described methods for constructing a large-scale anchor digital twin.

[0095] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] For a description of the computer-readable storage medium provided in this application, please refer to the above method embodiments; further details will not be repeated here.

[0097] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements 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 constructing a digital twin of an ultra-large anchorage, characterized in that, Includes the following steps: Step S1. Set up a three-dimensional monitoring array of inclinometer tubes that are evenly distributed in the horizontal and vertical directions on the anchorage diaphragm wall structure, and transmit the collected displacement data to the field data processing center in real time through an information transmission device; Step S2. Using the 2D surface unfolding interpolation method, the three-dimensional structural features of the anchorage diaphragm wall are converted into two-dimensional coordinates. In the unfolded two-dimensional coordinate system, the two-dimensional plane inverse distance interpolation method is used to complete the grid data interpolation expansion, and the grid point displacement values ​​and their corresponding coordinate data are stored in the database in real time in matrix form. Step S3. Based on the structural data of the anchorage diaphragm wall recorded during the design and construction process, create a BIM model, optimize the model structure and compress the model file size, and then reduce the computer memory usage and rendering calls during model loading and rendering by repeating component instantiation to obtain a lightweight model; Step S4. Initialize the 3D scene and configure the orthographic camera to adapt to the large anchorage structure; Set ambient light and parallel light sources to enhance the sense of depth; import the lightweight model from step S3 and configure the 3D scene on the web; align the grid data stored in the database according to step S2 with the BIM model coordinate system; then process the grid data through the shader and complete the heat map rendering and visualization. Step S5. Integrate multi-source heterogeneous data, including real-time monitoring data from sensor networks, BIM model data, and historical construction data, and display the monitoring data in a fusion manner to provide a multi-dimensional visualization interface.

2. The method for constructing a digital twin of an ultra-large anchorage according to claim 1, characterized in that, In step S1, the three-dimensional monitoring array is formed by a three-dimensional monitoring array of at least 16 inclinometer tubes that are evenly distributed along the horizontal and vertical directions. The depth of the inclinometer tubes is consistent with the diaphragm wall structure. The guide groove of the inclinometer tube is perpendicular to the edge of the pit and points to the center of the pit. The sensor in the inclinometer tube collects data once per hour.

3. The method for constructing a digital twin of an ultra-large anchorage according to claim 2, characterized in that, The information transmission device in step S1 includes a distributed optical fiber sensor demodulator, an industrial-grade switch, and an optical fiber network.

4. A method for constructing a digital twin of an ultra-large anchorage according to any one of claims 1-3, characterized in that... The 2D surface unfolding interpolation method in step S2 includes: (1) Coordinate system establishment: Set the central axis of the diaphragm wall cylinder as the z-axis of the Cartesian coordinate system, with the center of the cylinder ground as the origin; (2) Coordinate transformation: Replace the Cartesian sensor coordinates and values ​​(x,y,z) with cylindrical coordinates (r,θ,z); where r is the distance from the sensor to the central axis, θ is the angle of the sensor on the horizontal plane, θ∈[0,2π); (3) Boundary condition continuity processing: To ensure continuity at θ = 0 and θ = 2π, that is, the continuity of the left and right boundaries after the cylinder is unfolded, the data at θ = 0 is copied at θ = 2π. (4) Coordinate normalization: The angle θ is normalized according to the actual arc length of the cylindrical surface, that is, x_θ=θ×r, the x-axis represents the distance along the circumference, and the unit is the same as the z-axis (height); (5) The calculation formula is as follows: (Using the two-dimensional inverse plane distance interpolation method) In the formula It is the estimated value of position x0, Z(x i ,y i ) is a known position x i The observed values, and the predicted location of the unknown point (x0, y0). It is all known points (x) i y i The weighted sum of the values ​​of Z; the weight of each known point is determined by the reciprocal of its distance di from the unknown point to the power of α; di represents the straight-line distance between the unknown point and the i-th known point.

5. A method for constructing a digital twin of an ultra-large anchorage according to any one of claims 1-3, characterized in that, The specific methods for optimizing the model structure in step S3 include: deleting unnecessary internal components, removing hidden surfaces and overlapping geometry, merging coplanar small patches, simplifying complex surfaces, reducing visually insignificant areas of surface, organizing the model tree structure, and deleting redundant groups and subgroups to flatten the model structure.

6. The method for constructing a digital twin of an ultra-large anchorage according to claim 5, characterized in that, The specific methods for compressing the model file size in step S3 include: using the Draco compression algorithm to quantize vertex coordinates, converting floating-point numbers to integers to reduce data precision while maintaining visual quality; using predictive coding techniques to compress vertex data, utilizing the spatial relationships between adjacent vertices to reduce data redundancy; encoding the topology to reduce redundant information; and setting compression levels to complete model size compression while maintaining decoding speed.

7. A method for constructing a digital twin of an ultra-large anchorage according to any one of claims 1-3, characterized in that, The specific methods for the shader to process data in step S4 include: (1) Data normalization processing: Calculate the spatial coordinates and the given normalized displacement value v. Assume the minimum displacement value v_min and the maximum displacement value v_max. The normalization formula is: t = (v - v_min) / (v_max - v_min) (2) Using an additive mixing approach, for each pixel in the overlapping region, a new color value C_result is assigned: C_result=min(C1+C2+...+Cn,1.0) Where C1+C2+...+Cn are the color values ​​of the n overlapping points at the pixel position, and the result is compared with 1.0, taking the smaller value between the two.

8. A method for constructing a digital twin of an ultra-large anchorage according to any one of claims 1-3, characterized in that, Step S5 provides multi-dimensional visualization interface functions, including: providing multi-dimensional visualization interface, including 3D model view, 2D profile view and time series data chart.

9. A digital twin construction system for ultra-large anchorages, characterized in that: include: The large anchorage structure real-time data acquisition unit is used to set up a three-dimensional monitoring array of inclinometer tubes that are uniformly distributed in the horizontal and vertical directions on the anchorage diaphragm wall structure, and to transmit the acquired displacement data to the field data processing center in real time through an information transmission device. The mesh data generation unit is used to convert the three-dimensional structural features of the anchorage diaphragm wall into two-dimensional coordinates using a 2D surface unfolding interpolation method. In the unfolded two-dimensional coordinate system, a two-dimensional plane inverse distance interpolation method is used to complete the mesh data interpolation expansion, and the mesh point displacement values ​​and their corresponding coordinate data are stored in the database in real time in matrix form. The BIM lightweight processing unit is used to create a BIM model based on the structural data of the anchorage diaphragm wall recorded during design and construction, optimize the model structure, compress the model file size, and reduce computer memory usage and rendering calls during model loading and rendering by using repeated component instantiation, resulting in a lightweight model. The 3D point cloud real-time rendering unit is used to initialize the 3D scene and configure the orthogonal camera to adapt to the large structure of the anchorage. Use ambient light and parallel light sources to enhance the sense of depth; Import the lightweight model from step S3 and configure the 3D scene on the web. Align the grid data stored in the database according to step S2 with the BIM model coordinate system. Then process the grid data through shaders and complete the heat map rendering and visualization. Also, a digital twin platform unit is used to integrate multi-source heterogeneous data, including real-time monitoring data from sensor networks, BIM model data, and historical construction data, to perform fusion display of monitoring data and provide a multi-dimensional visualization interface.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor using the method for constructing a super-large anchor digital twin as described in any one of claims 1-8.

Images