A method, system, hardware and application for constructing a three-dimensional digital sand table for unmanned aerial vehicle real scene and engineering BIM simulation

Abstract

The application relates to a UAV real scene and engineering BIM simulation three-dimensional digital sand table construction method, system, hardware and application, obtains synthetic aperture radar image data of a target area, fuses D-InSAR technology and PS-InSAR technology processing, generates a wide-area ground surface deformation field map and time-series subsidence data, identifies and marks at least one potential risk area, plans a UAV earth-simulating flight route, controls a camera carried by the UAV to collect multi-angle image data of the potential risk area along the route, pre-processes, and generates a three-dimensional real scene model with geographic coordinate information; the BIM model of the target engineering and the three-dimensional real scene model are fused in a unified coordinate system to construct a three-dimensional digital sand table; based on the wide-area ground surface deformation field map, the engineering component parameters in the BIM model are dynamically adjusted, and the engineering construction process or the risk early warning is simulated in the three-dimensional digital sand table. The application effectively identifies the potential risk area and improves the digital level of engineering management.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, and in particular to a method, system, hardware, and application for constructing a three-dimensional digital sand table of UAV real-scene and engineering BIM simulation that combines engineering informatization and digital twin technology. Background Technology

[0002] With the rapid development of infrastructure construction, the scale of highway, railway, water conservancy, and municipal engineering projects is constantly expanding, and the construction environment is becoming increasingly complex, placing higher demands on the refined management and safety risk prevention and control of the entire project lifecycle. Traditional engineering monitoring and management methods mainly rely on manual inspections, total station measurements, and two-dimensional drawing reviews, which have problems such as limited monitoring scope, lagging data updates, serious information silos, and insufficient decision support capabilities, making it difficult to meet the needs of modern engineering for digital, visual, and intelligent management.

[0003] In the field of reality modeling, UAV oblique photogrammetry has become an important means of obtaining high-precision 3D reality models. However, existing UAV aerial surveying methods have significant shortcomings in complex terrain conditions: traditional flight path planning uses constant flight altitude, resulting in uneven image resolution and overlap in areas with large terrain undulations, affecting the accuracy of 3D reconstruction; the placement of ground control points lacks a systematic optimization scheme, making it difficult to balance control accuracy and operational efficiency; during multi-view image processing, the geometric accuracy control of bundle triangulation is insufficient, leading to local deformation or texture misalignment in the generated 3D model. In addition, existing UAV aerial surveying results are mostly single static models, lacking a mechanism for integrating with dynamic engineering information, making it difficult to support dynamic monitoring of the construction process.

[0004] In terms of engineering information modeling applications, Building Information Modeling (BIM) technology has been widely used in engineering design and construction management. However, existing BIM applications are mostly limited to the design stage or used independently, lacking effective integration with real-world models and monitoring data. Specifically, the coordinate systems of BIM models and real-world models are not unified, making it difficult to achieve accurate overlay and comparison. Monitoring data cannot drive dynamic adjustment of BIM model parameters, the model cannot reflect the real-time status of engineering entities, and the interactive functions of 3D visualization platforms are limited, making it difficult to support multi-person collaborative decision-making and immersive experiences.

[0005] Existing technologies have attempted to combine UAV reality models with BIM models to construct 3D digital sand tables, but they generally suffer from the following technical shortcomings: deformation monitoring is separated from reality modeling, making it impossible to use wide-area monitoring results as risk guidance for precise UAV aerial surveys; the integration of reality models and BIM models lacks standardized data conversion and coordinate unification processes, making it difficult to guarantee integration accuracy; the 3D digital sand table has limited functionality, lacking the integration of diverse algorithms such as path planning, heat map analysis, collision detection, and gesture interaction; and its application scenarios are limited to the construction phase, failing to cover foundation assessment during the construction preparation phase and structural health monitoring during the operation and maintenance phase. Summary of the Invention

[0006] This invention solves the problems existing in the prior art and provides a method, system, hardware and application for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation.

[0007] The technical solution adopted in this invention is a method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation, comprising the following steps: S1 acquires synthetic aperture radar image data of the target area, integrates D-InSAR and PS-InSAR technologies for processing, generates a wide-area surface deformation field map and time-series subsidence data, and identifies and marks at least one potential risk area. S2 Based on the terrain elevation data of the potential risk area, plan the drone's terrain-following flight path and control the camera on the drone to collect multi-angle image data of the potential risk area along the flight path; S3 preprocesses the multi-angle image data to generate a three-dimensional real-scene model with geographic coordinate information; S4 integrates the BIM model of the target project with the three-dimensional real-scene model in a unified coordinate system to construct a three-dimensional digital sand table; based on the wide-area surface deformation field map, it drives the dynamic adjustment of the engineering component parameters in the BIM model, and simulates the engineering construction process or warns of risks in the three-dimensional digital sand table.

[0008] Preferably, in S1, the two-track D-InSAR technology is used to remove the terrain phase using an external digital elevation model to generate a deformation field covering the entire target area; PS points are selected based on the amplitude deviation index and the temporal coherence coefficient dual thresholds, and the temporal deformation information of each PS point is calculated using PS-InSAR technology. The results identified by D-InSAR technology are spatially overlaid with the points located by PS-InSAR technology to form the deformation monitoring results.

[0009] Preferably, in S2, a digital elevation model of the target area is obtained or a digital surface model is obtained through pre-scan flight as the elevation basis; The optimal flight direction is determined based on the convex polygon width calculation method; according to the optimal flight direction, combined with the preset forward overlap, lateral overlap and photographic baseline length, a variable altitude flight path that varies with the terrain undulation is generated, so that the UAV maintains a relatively constant flight altitude with the ground during flight; the planning of the terrain-following flight path is completed.

[0010] Preferably, in S3, the preprocessing includes multi-view image dense matching, bundled aerial triangulation, and texture mapping; Aerial triangulation using the bundle method introduces image control point optimization. The image control points are arranged in an optimized layout, forming a triangle at the center of the survey area and sparsely distributed at the edges. The absolute error, mean absolute error, and standard deviation of the three-dimensional coordinates of the image control points are calculated to evaluate and correct the geometric accuracy of the three-dimensional reality model.

[0011] Preferably, in S4, the 3D digital sand table integrates one or more of the following interactive algorithms: A path planning module based on Dijkstra's algorithm or A* algorithm is used to simulate traffic flow. A heatmap generation module based on kernel density estimation algorithm is used to display data distribution density; The collision detection and dynamic simulation module based on the physics engine is used to simulate object movement or disaster scenarios. An interactive response module based on gesture recognition algorithms is used to realize natural human-computer interaction in the sand table.

[0012] Preferably, in S4, the 3D reality model is converted into FBX format and its geographic coordinate metadata is retained; the BIM model and the 3D reality model are unified to a standard coordinate system for fusion; the fusion model is parametrically debugged and visualized using a rendering engine to generate a 4D construction model containing construction progress information.

[0013] A UAV-based real-world scenario and engineering BIM simulation 3D digital sand table system is provided to implement the aforementioned method for constructing a UAV-based real-world scenario and engineering BIM simulation 3D digital sand table; the system includes: The data acquisition layer is used to acquire synthetic aperture radar image data of the target area, identify and mark at least one potential risk area, plan the UAV's terrain-following flight path, and collect multi-angle image data of the potential risk area. The data processing layer is used to preprocess multi-angle image data, generate a three-dimensional real scene model with geographic coordinate information, fuse the BIM model of the target project with the three-dimensional real scene model in a unified coordinate system, construct a three-dimensional digital sand table, drive the dynamic adjustment of the engineering component parameters in the BIM model, and simulate the engineering construction process or warn of risks in the three-dimensional digital sand table. The application interaction layer is used to provide a visual display and interactive interface for the 3D digital sand table.

[0014] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation.

[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation.

[0016] An application of the aforementioned method for constructing a 3D digital sand table using UAV real-scene and engineering BIM simulation, applicable to the entire engineering process; including: During the construction preparation phase, InSAR monitoring results were used to assess foundation stability and optimize route selection. During the construction phase, oblique photography by drones was integrated with BIM models to visualize and monitor the construction progress and calculate the earthwork volume. During the operation and maintenance phase, new reality models are generated based on updated UAV imagery and InSAR data and compared with BIM models to conduct structural deformation monitoring and health assessment.

[0017] This invention relates to a method, system, hardware, and application for constructing a 3D digital sand table using UAV real-scene and engineering BIM simulation. The method involves acquiring synthetic aperture radar image data of the target area, processing it using D-InSAR and PS-InSAR technologies to generate a wide-area surface deformation field map and time-series subsidence data, identifying and marking at least one potential risk area, planning a UAV-based terrain-following flight path, controlling the UAV's onboard camera to collect multi-angle image data of the potential risk area along the flight path, preprocessing the data, and generating a 3D real-scene model with geographic coordinate information. The BIM model of the target project is then fused with the 3D real-scene model in a unified coordinate system to construct a 3D digital sand table. Based on the wide-area surface deformation field map, the parameters of engineering components in the BIM model are dynamically adjusted, and the construction process or risk warning is simulated in the 3D digital sand table.

[0018] The beneficial effects of this invention are as follows: (1) By integrating D-InSAR and PS-InSAR technologies, multi-scale monitoring from large-scale deformation fields to point-like high-precision deformation can be achieved, effectively identifying potential risk areas and providing data support for engineering site selection and construction safety. (2) By planning the flight path and optimizing the layout of the image control points, the problem of uneven resolution of UAV images under complex terrain conditions is solved, and the generated three-dimensional real scene model has high geometric accuracy and realistic texture. (3) Integrate the BIM model with the real scene model in a unified coordinate system, and use InSAR deformation data to drive the dynamic adjustment of BIM component parameters to realize the visualization simulation and risk warning of the construction process, and improve the digital level of project management. (4) Integrating multiple interactive algorithms such as path planning, heat map generation, collision detection and gesture recognition enhances the usability and user experience of the 3D digital sand table and supports multi-person collaborative decision-making; (5) It covers the entire process of the project and provides a unified technical platform for the full life cycle management of the project. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the system structure of the present invention; Figure 3 This is a schematic diagram of the terrain-following flight path planning of the present invention; Figure 4 This is a schematic diagram of the ground-following flight of the present invention; Figure 5 This is a schematic diagram illustrating the fusion effect of the three-dimensional digital sand table according to the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] This invention relates to a method for constructing a 3D digital sand table of UAV real-scene and engineering BIM simulation. Based on multi-source data fusion and collaborative modeling technology, it constructs an integrated technical system of "monitoring-modeling-simulation-early warning". It utilizes synthetic aperture radar interferometry (SAR) technology to obtain large-scale surface subsidence trends in the construction area through D-InSAR, identifying potential risk zones. It integrates PS-InSAR technology and combines permanent scattering points for long-term series analysis to achieve millimeter-level deformation monitoring of key areas, providing early warning basis for construction safety. The UAV is equipped with a five-lens camera to simultaneously acquire high-resolution images from multiple vertical and oblique angles. Through collinearity equations and bundle adjustment, high-density point clouds are generated through dense matching of multi-view images, constructing a high-precision real-scene 3D model. The UAV real-scene model and the BIM parametric model are deeply fused in a unified coordinate system (WGS84), and the integration and sharing of geographic and engineering information are achieved through the Microstation platform. Parametric rendering and dynamic simulation are performed using the Blender platform, supporting construction progress tracking, scheme optimization, and risk early warning, forming a digital twin management platform that integrates "real-scene + information" multi-source data.

[0022] This method specifically includes the following steps: S1 acquires synthetic aperture radar image data of the target area, integrates D-InSAR and PS-InSAR technologies for processing, generates a wide-area surface deformation field map and time-series subsidence data, and identifies and marks at least one potential risk area. S2 Based on the terrain elevation data of the potential risk area, plan the drone's terrain-following flight path and control the camera on the drone to collect multi-angle image data of the potential risk area along the flight path; S3 preprocesses the multi-angle image data to generate a high-precision three-dimensional real-scene model with geographic coordinate information; S4 integrates the BIM model of the target project with the high-precision 3D reality model in a unified coordinate system to construct a 3D digital sand table; based on the wide-area surface deformation field map, it drives the dynamic adjustment of the engineering component parameters in the BIM model, and simulates the engineering construction process or warns of risks in the 3D digital sand table.

[0023] The following uses a mountainous expressway construction project as an example to illustrate the steps of this method in detail. The project is located in a reservoir section of a city in Zhejiang Province, with complex terrain and prominent mountain features, crisscrossed by hills and mountains, and a total route length of approximately 33.60 km.

[0024] S1 acquires synthetic aperture radar image data of the target area, integrates D-InSAR and PS-InSAR technologies for processing, generates a wide-area surface deformation field map and time-series subsidence data, and identifies and marks at least one potential risk area. In this invention, the main purpose of S1 is to integrate D-InSAR and PS-InSAR technologies to obtain wide-area deformation field and high-precision time-series settlement data of the target area, providing risk guidance for subsequent UAV-based refined aerial surveys.

[0025] In practice, synthetic aperture radar (SAR) image data covering the target area is acquired, Sentinel-1A satellite image with a suitable time baseline is selected, the imaging mode is interferometric wide swath mode (IW), the polarization mode is VV, and the data coverage swath width reaches 250km. (1-1) A two-track D-InSAR technique is employed, utilizing an external digital elevation model (DEM) to remove terrain phase and generate a deformation field covering the entire target area. Specifically, the two-track D-InSAR technique is used, where the two-track method refers to differential interferometry processing using two SAR images and an external DEM. An external 90m spatial resolution SRTM digital elevation model (DEM) is used to simulate the terrain phase, removing terrain influences. The acquired SAR images undergo precise orbit data encryption processing to generate interferograms and perform baseline estimation. Subsequently, through differential interferogram generation, filtering, and phase unwrapping steps, a wide-area surface deformation field map covering the entire target area is extracted; the interferometric phase φ... init Includes deformation phase φ def Terrain phase φ topo , reference ellipsoid phase φ refer Atmospheric delayed phase φ atmo and noise phase φ noise , satisfying φ init =φ def +φ topo +φ refer +φ atmo +φ noise After removing topography and reference ellipsoid phase through differential processing, the surface deformation d = -λ·φ is calculated. def / (4π), where λ is the radar wavelength.

[0026] (1-2) PS points are selected based on the dual thresholds of amplitude deviation index and temporal coherence coefficient, and high-precision temporal deformation information of each PS point is calculated using PS-InSAR technology. Based on D-InSAR processing, 12 time-series SAR images of the same area were selected. Based on the amplitude deviation index (ADI) and the temporal coherence coefficient, dual thresholds were set to screen out permanent scatterers (PS points) that remain stable in the long-term series. In this embodiment, the dual threshold conditions were set as ADI threshold less than 0.25 and the mean temporal coherence coefficient greater than 0.7. Through phase model calculation, considering the DEM error and atmospheric phase influence, high-precision temporal deformation information of each PS point was calculated.

[0027] (1-3) The results identified by D-InSAR technology and the points located by PS-InSAR technology are spatially overlaid and analyzed to form the deformation monitoring results.

[0028] The wide-area deformation field generated by D-InSAR is spatially superimposed with the deformation information of PS points located by PS-InSAR. Based on the settlement rate and deformation level, areas with a settlement rate greater than -15 mm / year or deformation exceeding a preset threshold are automatically marked as potential risk areas.

[0029] In practice, areas with a settlement rate greater than -15 mm / year, a cumulative deformation exceeding ±0.05 m, or a PS point density higher than a preset threshold are generally marked as potential risk areas.

[0030] S2 Based on the terrain elevation data of the potential risk area, plan the drone's terrain-following flight path and control the camera on the drone to collect multi-angle image data of the potential risk area along the flight path; In this invention, the multi-angle image data has the characteristics of high resolution and strong consistency.

[0031] S2 specifically includes the following steps: The digital elevation model (DEM) of the target area or the digital surface model (DSM) obtained through pre-scan flight is used as the elevation basis; in this embodiment, in order to ensure accuracy, a pre-scan flight is performed on the potential risk area to generate high-resolution DSM data; The optimal flight direction is determined based on a convex polygon width calculation method. Specifically, the vertices of the convex polygon representing the aerial survey area are arranged sequentially as V1, V2, V3...V... using this method. n ,like Figure 3 As shown, select edge V1V2, traverse all other points, calculate the distance from each point to V1V2, select the maximum distance as the width of the edge, traverse all edges in turn, sort the recorded edge widths, and select the edge direction corresponding to the minimum width as the optimal flight direction to minimize the number of flight path turns. Based on the optimal flight direction, combined with preset forward overlap, lateral overlap, and photographic baseline length, a variable altitude flight path that varies with terrain undulations is generated, ensuring that the UAV maintains a relatively constant flight altitude relative to the ground during flight. More precisely, based on the rate and extent of InSAR subsidence, the UAV's flight altitude, forward overlap, lateral overlap, and target resolution are dynamically calculated: the higher the subsidence rate and the larger the subsidence area, the lower the flight altitude, the higher the overlap, and the higher the resolution, thus completing the planning of the terrain-following flight path. For example, in one embodiment, when the subsidence rate v < 5 mm / year, the flight altitude is 150 m, the forward overlap is 70%, and the lateral overlap is 60%; when 5 ≤ v < 15 mm / year, the flight altitude is 100 m, and the overlap is 80% / 70%; when the subsidence rate c ≥ 15 mm / year, the flight altitude is 70 m, and the overlap is 85% / 75%. The larger the subsidence area, the lower the flight altitude or the higher the overlap. Calculate the absolute flight altitude H for each waypoint based on the DSM elevation values. abs =H ground +H relative H ground H represents the elevation of a ground point. relative To preset a relative altitude, such as 100m, so that the drone maintains a constant altitude relative to the ground during flight, such as... Figure 4 As shown, the optimal flight path map covering the entire survey area is generated. Clearly, when the target area is an urban road network, the UAV flight path planning is limited by building height and airspace, requiring the addition of obstacle avoidance constraints. In this case, the correction formula is H... abs =max(H ground +H relative H building +H safety ), where H building H represents the building height. safety A preset safe distance is required. In practice, the so-called terrain-following flight refers to the UAV adjusting its flight altitude in real time according to the terrain undulations during flight, so that the flight altitude remains constant relative to the ground, which is different from traditional altitude-contact flight.

[0032] After completing the above processing, control the multi-rotor UAV equipped with the camera to fly along the planned terrain-following route. During the flight, the camera's five lenses (one vertical and four tilted) are exposed simultaneously to collect multi-angle image data of potential risk areas, ensuring that the images have both high resolution and rich information on the side texture of ground features.

[0033] S3 preprocesses the multi-angle image data to generate a high-precision three-dimensional real-scene model with geographic coordinate information; Preprocessing includes multi-view image dense matching, bundled aerial triangulation, and texture mapping; Specifically, the acquired images are first radiometrically corrected to eliminate the effects of uneven illumination. Then, distortion correction is performed on the images using camera calibration parameters to correct radial distortion, tangential distortion, and CCD array deformation caused by lens optical defects. The radial distortion correction formula is as follows: △x=x(K1*r 2 +K2*r 4 +K3*r 6 ) △y = y(K1*r 2 +K2*r 4 +K3*r 6 ) Where r is the radial direction, and K1, K2, and K3 are radial distortion parameters.

[0034] In multi-view image dense matching, a feature point extraction and matching method based on the SIFT algorithm is adopted. By constructing a scale space, detecting extreme points, locating feature points, assigning directions, and generating descriptors, the scale space of the image is constructed and stable feature points are extracted. Extreme points are detected by constructing Gaussian pyramids and difference-of-Gaussian pyramids (DOG), key points are accurately located and unstable edge response points are removed. The principal gradient direction of feature points is calculated to generate a 128-dimensional feature vector descriptor, realizing fast feature point matching. The aerial triangulation method using bundled beams introduces image control point optimization. The image control points are arranged in an optimized layout, forming a triangular pattern at the center of the survey area and sparsely distributed at the edges. In practice, the triangular layout refers to selecting three non-collinear image control points at the center of the survey area to form a triangular network. The interior angles of the triangular network are between 30° and 150°. Generally, the distance between each point is not less than 50m, and the distance between image control points at the edges can be increased to more than 100m. Specifically, control points are mainly deployed at the center of the test area, with adjacent control points forming a triangular relationship as much as possible, and are placed in prominent locations such as road corners. Only a small number of control points are deployed at the edge of the test area. The absolute error AE, mean absolute error MAE, and standard deviation σ of the three-dimensional coordinates of the control points are calculated to evaluate and correct the geometric accuracy of the three-dimensional reality model. Here, AE is the difference between the measured value and the true value, MAE is the arithmetic mean of the absolute errors of each point, and σ reflects the dispersion of the error. The calculation methods of AE, MAE, and σ are well-known techniques in the field. Calculations show that the MAE of each control point in the X, Y, and Z directions is better than 0.03m, and the σ is better than 0.04m, with the highest accuracy in the Z direction (MAE better than 0.003m). The above accuracy indicators meet the specifications for 1:500 large-scale topographic mapping, indicating that the optimized placement method of control points and bundle-based aerial triangulation can effectively correct the geometric accuracy of the three-dimensional reality model.

[0035] An irregular triangular mesh is constructed based on the solved high-density point cloud to form a 3D white model. A two-step mapping method is used to map the image texture in the texture space onto the surface of the 3D white model, generating a high-precision 3D real scene model with realistic visual effects, which is then stored in OSGB or FBX format.

[0036] When dealing with urban road networks, it is also necessary to increase the finer modeling of road markings and traffic facilities, and ensure the accuracy of road surface defects (cracks, potholes) identification by setting higher texture resolution and denser point cloud density.

[0037] S4 integrates the BIM model of the target project with the high-precision 3D reality model in a unified coordinate system to construct a 3D digital sand table; based on the wide-area surface deformation field map, it drives the dynamic adjustment of the engineering component parameters in the BIM model, and simulates the engineering construction process or warns of risks in the 3D digital sand table.

[0038] The 3D reality model is converted to FBX format while retaining its geographic coordinate metadata; the BIM model and the 3D reality model are unified and merged under a standard coordinate system; the merged model is parametrically debugged and visualized using a rendering engine to generate a 4D construction model containing construction progress information.

[0039] Specifically, the generated 3D reality model is converted into FBX format, and its geographic coordinate metadata is retained. The master file is placed in the data directory, and the file recording the geographic location information of the model is placed in the same folder as the master file. Microstation is used as the fusion platform to unify the BIM model of the target project and the 3D reality model into the WGS84 coordinate system for fusion, so as to achieve accurate overlay of the BIM model and the real terrain.

[0040] In this embodiment, the BIM model of the target project can be constructed using industry-standard BIM modeling software, including but not limited to Revit, Open Bridge Modeler, and Microstation. The BIM model contains parametric information such as the geometric information, material properties, and construction schedule of the engineering components, and is exported in open formats such as IFC and DNG for subsequent integration with the real-world model. Taking bridge engineering as an example, the BIM model contains the three-dimensional geometric information and associated attributes of components such as main beams, piers, and abutments; taking roadbed engineering as an example, the BIM model contains construction parameters such as fill layer thickness, slope gradient, and compaction degree. The specific creation method of the BIM model adopts known techniques in the field and will not be elaborated here. Based on the wide-area surface deformation field map obtained by S1, the settlement / lift values ​​of each engineering station or key component are extracted. Using the parametric modeling function of a general-purpose 3D rendering engine (such as Blender, Unreal Engine, Unity, etc.), these deformation data are used as driving parameters to dynamically adjust component parameters such as roadbed fill height and slope gradient in the BIM model. Specifically, the settlement / uplift values ​​are directly superimposed onto the original design elevation; that is, the filling height is reduced by the same amount of settlement, and increased by the same amount of uplift. Of course, adjustment coefficients can also be configured in practical applications, which is easily understood by those skilled in the art. When the settlement rate of a certain embankment section exceeds the warning value, the system automatically adjusts the design elevation of that section of the roadbed in the BIM model and simulates the subsequent earthwork allocation plan in a 3D digital sand table, providing early warning of potential landslide risks. This means that the system automatically corrects the roadbed filling parameters or warning thresholds based on updates from InSAR and drones.

[0041] The 3D digital sand table integrates one or more of the following interactive algorithms: A path planning module based on Dijkstra's algorithm or A* algorithm is used to simulate traffic flow, that is, to simulate the transportation routes of construction vehicles on three-dimensional terrain and optimize material scheduling. The heat map generation module based on the kernel density estimation (KDE) algorithm, more precisely, generates heat maps of the settlement rate distribution monitored by PS-InSAR to show the data distribution density and highlight high-risk areas. The collision detection and dynamic simulation module based on the physics engine is used to simulate object movement or disaster scenarios. The former mainly includes simulating the collision risk during tower crane rotation or component hoisting. The interactive response module, based on a gesture recognition algorithm (convolutional neural network), is used to realize natural human-computer interaction of the sand table. Specifically, the camera captures gesture images, the CNN extracts features, recognizes gestures such as single-finger click, two-finger zoom, and three-finger rotation, and sends control commands to realize zooming, rotation and object picking operations on the three-dimensional digital sand table.

[0042] This invention also relates to a UAV real-scene and engineering BIM simulation 3D digital sand table system, used to implement the aforementioned UAV real-scene and engineering BIM simulation 3D digital sand table construction method; the system includes: (1) Data acquisition layer, used to acquire synthetic aperture radar image data of the target area, identify and mark at least one potential risk area, plan the UAV's terrain-following flight path, and acquire multi-angle image data of the potential risk area; specifically, including: The InSAR data receiving unit is used to acquire SAR images and run an algorithm that fuses D-InSAR and PS-InSAR to generate deformation monitoring maps. The flight path planning unit generates UAV flight paths based on deformation monitoring results and elevation data using a convex polygon width algorithm and a terrain-following flight algorithm. The UAV flight control unit communicates with the flight path planning unit to control the UAV equipped with multi-view cameras to perform flight missions and transmit flight attitude and image data back in real time.

[0043] (2) Data processing layer, used to preprocess multi-angle image data, generate a high-precision 3D reality model with geographic coordinate information, fuse the BIM model of the target project with the high-precision 3D reality model in a unified coordinate system, construct a 3D digital sand table, drive the dynamic adjustment of engineering component parameters in the BIM model, and simulate the engineering construction process or warn of risks in the 3D digital sand table; specifically, including: The 3D reconstruction unit has a built-in 3D reconstruction engine, which is used to perform aerial triangulation, dense matching, and texture mapping on multi-angle images to generate 3D reality models. The model fusion unit is used to perform coordinate transformation, format unification, and spatial alignment between the BIM model and the 3D reality model; The dynamic simulation unit is used to receive InSAR deformation data and drive the dynamic adjustment of BIM model parameters to simulate the construction process and risk scenarios.

[0044] (3) Application interaction layer, used to provide a visual display and interactive operation interface for the 3D digital sand table; specifically, including: The visualization rendering module, based on the Blender rendering engine, provides glasses-free 3D effects and a technologically advanced interface; The interactive control module integrates algorithm interfaces such as path planning, heat map generation, collision detection, and gesture recognition, and supports users to operate via multi-touch or mouse and keyboard. The business integration module provides API interfaces for integrating IoT data (such as video surveillance, BeiDou positioning, and environmental sensors). User business systems (such as project management and schedule planning systems) support interface customization and secondary development.

[0045] The present invention also relates to an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for constructing a three-dimensional digital sand table of UAV real-scene and engineering BIM simulation.

[0046] The present invention also relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for constructing a three-dimensional digital sand table of UAV real-scene and engineering BIM simulation.

[0047] This invention also relates to the application of the aforementioned method for constructing a three-dimensional digital sand table using UAV real-scene and engineering BIM simulation, applicable to the entire engineering process; including: During the construction preparation phase, InSAR monitoring results were used to assess the foundation stability and optimize the route selection. Deformation monitoring results generated by S1 were used to assess the foundation stability of the proposed route. Based on this, the design team adjusted the route selection plan to avoid high-risk areas and carried out pre-reinforcement design for special roadbed sections that could not be avoided, thus mitigating engineering risks from the source.

[0048] During the construction phase, oblique photography by drones is integrated with BIM models for visualized monitoring of construction progress and calculation of earthwork volume. In particular, during the subgrade filling and bridge pile foundation construction, the drone-BIM fusion mechanism of S2~S4 is used for visualized monitoring of construction progress. By periodically collecting drone images, new real-scene models are generated and compared with the planned progress in the BIM model to automatically calculate the difference in earthwork filling volume and identify lagging processes. When it is found that the actual settlement of a certain section of the subgrade is greater than the BIM simulation value, the system automatically pushes early warning information and re-simulates the subsequent construction plan.

[0049] During the operation and maintenance phase, a new reality model is generated based on updated UAV imagery and InSAR data and compared with the BIM model to conduct structural deformation monitoring and health assessment. After project completion, the final reality model and BIM model are used as the operation and maintenance base map. UAVs are used to collect images of the entire line every year, and combined with InSAR data, regular structural deformation monitoring is carried out. The newly generated reality model is compared with the operation and maintenance base map to automatically identify defects such as cracks, settlement, and slope collapse, quantify the rate of defect development, and provide accurate data support for maintenance decisions.

[0050] 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, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0051] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0052] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0053] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0054] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for constructing a 3D digital sand table for drone-based real-world scenarios and engineering BIM simulation, characterized in that, Includes the following steps: S1 acquires synthetic aperture radar image data of the target area, integrates D-InSAR and PS-InSAR technologies for processing, generates a wide-area surface deformation field map and time-series subsidence data, and identifies and marks at least one potential risk area. S2 Based on the terrain elevation data of the potential risk area, plan the drone's terrain-following flight path and control the camera on the drone to collect multi-angle image data of the potential risk area along the flight path; S3 preprocesses the multi-angle image data to generate a three-dimensional real-scene model with geographic coordinate information; S4 integrates the BIM model of the target project with the three-dimensional real-scene model in a unified coordinate system to construct a three-dimensional digital sand table; based on the wide-area surface deformation field map, it drives the dynamic adjustment of the engineering component parameters in the BIM model, and simulates the engineering construction process or warns of risks in the three-dimensional digital sand table.

2. The method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation according to claim 1, characterized in that, In S1, the two-track D-InSAR technology is used to remove the terrain phase using an external digital elevation model to generate a deformation field covering the entire target area; PS points are selected based on the amplitude deviation index and the temporal coherence coefficient dual thresholds, and the temporal deformation information of each PS point is calculated using PS-InSAR technology. The results identified by D-InSAR technology are spatially overlaid with the points located by PS-InSAR technology to form the deformation monitoring results.

3. The method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation according to claim 1, characterized in that, In S2, a digital elevation model of the target area is obtained or a digital surface model is obtained through pre-scan flight as the elevation basis; The optimal flight direction is determined based on the convex polygon width calculation method. Based on the optimal flight direction, combined with the preset forward overlap, lateral overlap and photographic baseline length, a variable altitude flight path that varies with the terrain undulation is generated so that the UAV maintains a relatively constant flight altitude with the ground during flight. Complete the planning of the terrain-following flight route.

4. The method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation according to claim 1, characterized in that, In S3, preprocessing includes multi-view image dense matching, bundled aerial triangulation, and texture mapping; Aerial triangulation using the bundle method introduces image control point optimization. The image control points are arranged in an optimized layout, forming a triangle at the center of the survey area and sparsely distributed at the edges. The absolute error, mean absolute error, and standard deviation of the three-dimensional coordinates of the image control points are calculated to evaluate and correct the geometric accuracy of the three-dimensional reality model.

5. The method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation according to claim 1, characterized in that, In S4, the 3D digital sand table integrates one or more of the following interactive algorithms: A path planning module based on Dijkstra's algorithm or A* algorithm is used to simulate traffic flow. A heatmap generation module based on kernel density estimation algorithm is used to display data distribution density; The collision detection and dynamic simulation module based on the physics engine is used to simulate object movement or disaster scenarios. An interactive response module based on gesture recognition algorithms is used to realize natural human-computer interaction in the sand table.

6. The method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation according to claim 1, characterized in that, In S4, the 3D reality model is converted to FBX format while retaining its geographic coordinate metadata; the BIM model and the 3D reality model are unified to a standard coordinate system for fusion; the rendering engine is used to parametrically debug and visualize the fused model to generate a 4D construction model containing construction progress information.

7. A three-dimensional digital sand table system for drone-based real-world scene and engineering BIM simulation, characterized in that, A method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation as described in any one of claims 1 to 6; the system includes: The data acquisition layer is used to acquire synthetic aperture radar image data of the target area, identify and mark at least one potential risk area, plan the UAV's terrain-following flight path, and collect multi-angle image data of the potential risk area. The data processing layer is used to preprocess multi-angle image data, generate a three-dimensional real scene model with geographic coordinate information, fuse the BIM model of the target project with the three-dimensional real scene model in a unified coordinate system, construct a three-dimensional digital sand table, drive the dynamic adjustment of the engineering component parameters in the BIM model, and simulate the engineering construction process or warn of risks in the three-dimensional digital sand table. The application interaction layer is used to provide a visual display and interactive interface for the 3D digital sand table.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation as described in any one of claims 1 to 6.

10. An application of the method for constructing a three-dimensional digital sand table for UAV real-scene and engineering BIM simulation as described in any one of claims 1 to 6, characterized in that, Applied to the entire engineering process; including: During the construction preparation phase, InSAR monitoring results were used to assess foundation stability and optimize route selection. During the construction phase, oblique photography by drones was integrated with BIM models to visualize and monitor the construction progress and calculate the earthwork volume. During the operation and maintenance phase, new reality models are generated based on updated UAV imagery and InSAR data and compared with BIM models to conduct structural deformation monitoring and health assessment.

Images