High-precision positioning and measuring method and system for offshore photovoltaic power station pile foundation

Abstract

The invention discloses a high-precision positioning and measuring method and system for a pile foundation of an offshore photovoltaic power station, and relates to the technical field of offshore photovoltaic engineering survey, and the high-precision positioning and measuring method for the pile foundation of the offshore photovoltaic power station mainly comprises the steps: generating a main route and a cross validation route according to engineering drawing coordinates of a constructed pile foundation by using a particle swarm optimization algorithm; and the known coordinates of the multiple ground control points are used for correcting the coordinates of the unmanned aerial vehicle to obtain coordinate conversion parameters, camera calibration, ortho-rectification and target detection are conducted on the target image to obtain the absolute coordinates of the center of the pile foundation, and the engineering drawing coordinates of the constructed pile foundation are compared to obtain the pile foundation construction error value. By implementing the high-precision positioning measurement method and system for the pile foundation of the offshore photovoltaic power station, the measurement precision and efficiency of the pile foundation of the offshore photovoltaic power station can be improved.

Description

Technical Field

[0001] This invention relates to the field of marine photovoltaic engineering measurement technology, and more specifically, to a high-precision positioning measurement method and system for marine photovoltaic power station pile foundations. Background Technology

[0002] In the construction of offshore photovoltaic projects, accurate measurement of pile foundation locations is a core prerequisite for ensuring the compliance of the overall layout of the photovoltaic array, the accuracy of component installation, and the stability of the engineering structure. Traditional pile foundation measurement methods mainly rely on manual measurement from a ship. This method is easily affected by complex environmental factors such as wind, waves, tides, and salt spray at sea. On the one hand, manual operation is greatly affected by the experience and physical strength of the operators and the visibility of the environment, resulting in low work efficiency. On the other hand, the measuring equipment is prone to reading deviations in turbulent environments.

[0003] While existing UAV surveying technology boasts advantages such as high efficiency and flexibility, a dedicated solution for measuring the location of offshore photovoltaic pile foundations has yet to be developed. Firstly, fluctuations in the UAV's flight attitude and errors in the lidar scanning angle can easily lead to distorted measurement data. Secondly, UAV cameras are affected by lens optical characteristics and shooting angles, resulting in radial and tangential distortions. Without targeted correction, these distortions directly impact the accuracy of pile foundation contour and coordinate extraction. Thirdly, the lack of a unified coordinate transformation system leads to significant conversion errors between the UAV's own coordinate system and the engineering coordinate system, making it difficult to output absolute pile foundation coordinates that meet engineering requirements.

[0004] Therefore, traditional manual shipborne surveying methods are greatly affected by the marine environment, have low efficiency, and are difficult to guarantee accuracy. Existing UAV surveying solutions have not yet formed a systematic solution in terms of coordinate system transformation and image distortion correction. As a result, the construction accuracy of existing technologies is difficult to meet the stringent construction requirements of ±5cm plane error and ±3cm elevation error for pile foundation positioning in offshore photovoltaic projects. There is an urgent need for a UAV surveying solution that is adapted to the complex marine environment and balances accuracy and efficiency. Summary of the Invention

[0005] The purpose of this invention is to provide a high-precision positioning and measurement method and system for offshore photovoltaic power station pile foundations, which can improve the measurement accuracy and efficiency of offshore photovoltaic power station pile foundations.

[0006] This invention provides a high-precision positioning and measurement method for offshore photovoltaic power station pile foundations, comprising the following steps: S1: Based on the engineering coordinates of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters; S2: Obtain the target image according to the main flight path and cross-validation flight path; perform camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center; S3: Obtain the pile foundation construction error value based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation.

[0007] The present invention also provides a high-precision positioning and measurement system for offshore photovoltaic power station pile foundations, the system comprising: Flight path planning module: Based on the coordinates of the engineering drawings of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters; Point cloud preprocessing, pile foundation feature and coordinate correction module: acquires target image according to the main flight path and cross-validation flight path; performs camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center; Accuracy assessment module: Based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation, the construction error value of the pile foundation is obtained.

[0008] The high-precision positioning and measurement method and system for offshore photovoltaic power station pile foundations provided by this invention has the following beneficial effects: This invention establishes ground control points and acquires precise RTK coordinates, which are simultaneously scanned by the UAV during measurement tasks. Based on the known coordinates of the control points and the coordinates scanned by the LiDAR, the real-time positioning information of the pile foundation is extracted through the UAV map management system and compared with the pile foundation design coordinates in the CAD engineering drawings. Construction errors are automatically calculated and annotated. Through multi-flight design and data fusion, the UAV's own errors are effectively eliminated, improving measurement accuracy. A coordinate correction method based on ground control points solves the problem of UAV positioning deviation. The system has a high degree of integration, automating the entire process from flight path planning and data acquisition to processing and analysis. It provides a direct and reliable reference for construction personnel to accurately adjust the pile foundation extension plate, improving measurement efficiency and reducing the risks of marine operations. The measurement accuracy meets the requirements for offshore photovoltaic pile foundation construction, ensuring the precise installation of photovoltaic arrays and reducing subsequent adjustment costs.

[0009] Specifically, this invention represents a breakthrough upgrade over traditional measurement schemes in terms of technological advancement. Through a hardware combination of "high-precision lidar + dual-frequency GNSS module," coupled with a flight path design where "the main flight path follows the array direction, and the cross-verification flight paths are at a 45° angle with varying altitudes," it effectively offsets the fluctuations in UAV flight attitude and the errors in lidar scanning angles, solving the pain point of insufficient accuracy in existing UAV solutions. Simultaneously, it constructs a fully automated system encompassing "CAD coordinate import → automatic flight path generation → real-time data acquisition → coordinate correction → error BIM overlay display," not only controlling measurement accuracy within ±5cm plane error and ±3cm elevation error, but also significantly surpassing the efficiency of traditional manual shipborne measurement. It completely eliminates the interference of sea waves and tides on manual operation, precisely matching the stringent construction requirements of offshore photovoltaic pile foundations.

[0010] This invention significantly reduces the measurement costs of offshore photovoltaic projects from a full life-cycle perspective. On the one hand, it replaces traditional manual shipboard measurement with automated processes, reducing a significant amount of manpower and time spent at sea, while also mitigating safety risks for personnel in complex marine environments, indirectly reducing safety protection and manpower management costs. On the other hand, it achieves precise positioning of the pile foundation center using YOLOv5 target detection and semantic segmentation technology, and automatically marks construction errors using a CAD engineering BIM overlay system, enabling early detection of pile foundation offset issues, avoiding large-scale rework during the later installation of photovoltaic modules, and reducing secondary adjustment costs due to insufficient accuracy, thus forming a complete cost control chain from early measurement to later construction.

[0011] The technical method of this invention has rigorous scientific theoretical support and quantifiable implementation logic. Addressing the image distortion problem, it derives a method including radial distortion through mathematical modeling. , , Coefficient) and tangential distortion ( , The formula for merging coefficients was used, and 30 images from different angles were collected using a checkerboard calibration target to accurately obtain parameters such as lens focal length and aperture, ensuring that the distortion correction process is traceable and reproducible. In the coordinate transformation stage, a four-layer transformation logic of "pixel coordinate system → camera coordinate system → UAV body coordinate system → geographic coordinate system" was strictly followed, with each layer relying on corresponding hardware or algorithms to provide data support. At the same time, an accuracy evaluation system of "point cloud denoising → feature point extraction → coordinate comparison → error source tracing" was established. When the error exceeds the limit, the solution can be optimized in reverse, rather than relying on experience judgment, which fully reflects the rigor and objectivity of scientific research.

[0012] This invention possesses strong potential for widespread adoption. On one hand, it has broad applicability in various application scenarios. Its core technologies, such as "RTK calibration of ground control points," "multi-route data fusion," and "image distortion correction," are not only suitable for offshore photovoltaic pile foundation measurement but can also be seamlessly extended to similar offshore engineering measurement scenarios, such as offshore wind power pile foundations and cross-sea bridge foundations, without requiring significant adjustments to the core technology framework. On the other hand, it lowers the barrier to entry in terms of operation. Route planning supports a combination of manual adjustment and automatic optimization, and on-site operators can obtain preset routes via remote control without needing to master complex algorithms. The data processing results are presented in a BIM visualization format, allowing construction personnel to intuitively read error information and use it for pile foundation adjustments without requiring extensive professional training. Both its technical adaptability and ease of operation lay the foundation for large-scale promotion. Attached Figure Description

[0013] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart of the high-precision positioning and measurement method for offshore photovoltaic power station pile foundations provided by the present invention; Figure 2 This is a flowchart of the method for measuring the location of pile foundations using an unmanned aerial vehicle (UAV) provided by the present invention; Figure 3 This is the particle swarm optimization algorithm flow provided by the present invention; Figure 4 This is a schematic diagram of attitude error in orthogonal aerial photography provided by the present invention; Figure 5 This is a schematic diagram of the radial and tangential distortion of the camera provided by the present invention; Figure 6 This is a schematic diagram of the image digital differential correction process provided by the present invention; Figure 7 This is a schematic diagram of the calculation of the pile foundation center by the unmanned aerial vehicle system provided by the present invention; Figure 8 This is a communication diagram of the UAV data and management system provided by the present invention; Figure 9 This is a schematic diagram of the drone map system provided by the present invention. Detailed Implementation

[0014] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0015] Figure 1 A schematic diagram of the high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to this embodiment is shown. In this embodiment, the high-precision positioning and measurement method for offshore photovoltaic power station pile foundations includes the following steps: S1: Based on the engineering coordinates of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters.

[0016] In one exemplary embodiment, the objective function of the particle swarm optimization algorithm is: , , , in, The objective function is... , , Let be the weights, and their sum is ; , and These represent the shortest path objective function, the low energy consumption objective function, and the shortest time objective function, respectively. , and Let n be the horizontal x-coordinate, horizontal y-coordinate, and altitude of the UAV; where n is the number of waypoints. The speed energy consumption coefficient, The energy consumption coefficient is the energy consumption coefficient based on altitude difference. Based on energy consumption, For the first The flight speed of the segment of the route.

[0017] In one exemplary embodiment, the constraints of the particle swarm optimization algorithm include endurance constraints, obstacle avoidance constraints, and task constraints; the endurance constraint is: total route distance. The maximum flight range of the drone is used as a reserve for emergency power; the obstacle avoidance constraint is the distance from all waypoints on the flight path to obstacles. Safety threshold; the task constraints are: if it is an inspection task, it covers a designated area, and the overlap rate between the route and the area is ≥95%; if it is a transportation task, it passes through the take-off and landing points and the target point.

[0018] In one exemplary embodiment, the particle encoding of the particle swarm optimization algorithm adopts waypoint sequence encoding, with each particle directly corresponding to a set of waypoint coordinates; when updating waypoint coordinates, the x and y axes are adjusted by no more than 100 meters per iteration step, and the z axis by no more than 50 meters; the particle number of the particle swarm optimization algorithm is 150, the inertia weight is 0.7, and the learning factor is 1.7.

[0019] In one exemplary embodiment, the velocity update formula of the particle swarm optimization algorithm is:

[0020] in, and Let represent the velocity of the i-th particle in the j-th dimension at generation t and generation (t+1); Indicates inertia weight; and Indicates the learning factor; and Represents a random number within the interval [0,1], enhancing the randomness of the algorithm's search; This represents the historical best position of the i-th particle in the j-th dimension; and This represents the position of the i-th particle in the j-th dimension at generation t and generation (t+1); This represents the globally optimal position in the j-th dimension of the entire particle swarm.

[0021] The position update formula for the particle swarm optimization algorithm is: .

[0022] In one exemplary embodiment, the iteration termination condition of the particle swarm optimization algorithm is the maximum number of iterations, or the change in the globally optimal fitness value over 10 consecutive iterations. .

[0023] In one exemplary embodiment, the maximum number of iterations is 50.

[0024] In one exemplary embodiment, the main flight path is used to enable the UAV to fly along the planned row direction of the photovoltaic array, and the spacing between adjacent flight paths of the main flight path is equal to the effective scanning width of the UAV's lidar, in order to ensure that the pile foundation is within the scanning range.

[0025] The cross-validation route forms a 45° angle with the main route, and the flight altitudes of the main route and the cross-validation route are different; there are no fewer than two cross-validation routes, and all cross-validation routes intersect. The cross-validation route is used to eliminate changes in UAV flight attitude and errors in lidar scanning angle by fusing data from multiple routes.

[0026] In one exemplary embodiment, the method for obtaining the coordinate transformation parameters is as follows: using RTK to obtain the known coordinates of multiple ground control points, using the lidar of the UAV to obtain the lidar scan coordinates of the multiple ground control points; and using the known coordinates of the multiple ground control points and the lidar scan coordinates of the multiple ground control points to perform matching calculations to solve for the coordinate transformation parameters.

[0027] S2: Obtain the target image according to the main route and cross-validation route; perform camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center.

[0028] In one exemplary embodiment, the camera calibration process includes: Camera intrinsic parameters and distortion coefficients are obtained from multiple calibration images taken from different angles; The target image is corrected for image distortion using the camera intrinsic parameters and distortion coefficients to obtain a target image without lens error.

[0029] In one exemplary embodiment, the orthorectification process includes: A digital elevation model is obtained based on point cloud data; Based on the digital elevation model and the orthophoto output grid, the original image coordinates are solved in reverse for each pixel of the orthophoto using the collinearity equation. Based on the original image coordinates, grayscale resampling is performed using bilinear interpolation to obtain an orthophoto of the pile foundation with accurate geographic coordinates and no geometric distortion.

[0030] In one exemplary embodiment, the target detection process includes: Based on the orthophoto of the pile foundation, the YOLOv5 target detection model is used to locate the boundary box of the pile foundation. The image pixel coordinates of the center point are obtained by using the intersection of the diagonals of the pile foundation boundary frame; The image pixel coordinates are converted into camera coordinates using camera intrinsic parameters; The camera coordinates are converted into UAV body coordinates using attitude sensor parameters; The coordinates of the UAV body are converted into geographic coordinates using the positioning system coordinates to obtain the absolute coordinates of the pile foundation center.

[0031] As an exemplary embodiment, the positioning system coordinates are GNSS coordinates.

[0032] S3: Obtain the pile foundation construction error value based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation.

[0033] This embodiment provides a high-precision positioning and measurement system for offshore photovoltaic power station pile foundations, the system comprising: Flight path planning module: Based on the coordinates of the engineering drawings of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters; Point cloud preprocessing, pile foundation feature and coordinate correction module: acquires target image according to the main flight path and cross-validation flight path; performs camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center; Accuracy assessment module: Based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation, the construction error value of the pile foundation is obtained.

[0034] In some embodiments, the above-described high-precision positioning and measurement method for offshore photovoltaic power station pile foundations can also be implemented in the following ways.

[0035] In this embodiment, the UAV is equipped with a high-precision lidar and a dual-frequency GNSS module. The lidar is used to acquire three-dimensional point cloud data of the pile foundation and surrounding terrain, while the dual-frequency GNSS module provides centimeter-level positioning accuracy.

[0036] Drone route selection: Based on the layout of the photovoltaic field and the parameters of the lidar, the main route and cross-validation route are generated, and manual adjustment and automatic optimization are supported.

[0037] Main flight path: fly along the planned direction of the photovoltaic array, with the distance between adjacent flight paths equal to the effective scanning width of the lidar, to ensure that the pile foundation is within the scanning range.

[0038] Cross-validation flight path: Flying at a 45° angle to the main flight path and at a different altitude. Establish two or more cross-flight paths. Through data fusion of multiple flight paths, eliminate changes in UAV flight attitude and errors in LiDAR scanning angles.

[0039] UAV coordinate correction method: Multiple ground control points are deployed in the offshore photovoltaic field area, and their precise coordinates are obtained using RTK measurement. During UAV measurement, a lidar simultaneously scans the ground control points. During data processing, the known coordinates of the ground control points and the lidar scan coordinates are matched and calculated to solve for the coordinate transformation parameters. Correction of pile foundation measurement coordinates: The pile foundation lidar coordinates acquired by the UAV (based on the UAV's own coordinate system) are converted to the engineering coordinate system using the above transformation parameters, eliminating UAV positioning deviations and system errors.

[0040] Image post-processing methods are used for UAV camera calibration and correction: Image distortion caused by the camera mainly stems from lens optical characteristics and shooting angle, and is divided into radial distortion and tangential distortion. Radial distortion is caused by the lens shape, manifesting as a difference in magnification between the image center and edges, such as barrel or pincushion distortion. Tangential distortion occurs because the camera is not perpendicular to the target surface, causing the object's edges to tilt or deform. Both types of distortion can lead to distortion of key information such as pile foundation coordinates and contours. Due to factors such as lens focal length, aperture, and lens design, UAVs exhibit both radial and tangential distortion. Failure to perform image mirror correction on the captured images will directly affect the accuracy of subsequent measurements.

[0041] Radial distortion:

[0042] Tangential distortion:

[0043] Merging distortions:

[0044] Camera calibration algorithms can provide more accurate lens focal length, aperture, and lens design parameters.

[0045] Orthorectification is used to correct systematic distortions in aerial images: eliminating image distortions caused by factors such as sensor viewpoint, terrain undulation, and atmospheric refraction, thereby obtaining images with accurate geographic coordinates and no geometric distortion. The main orthorectification process is as follows: defining the output orthorectified image grid; inverse mapping calculation: for each pixel of the output orthorectified image... Elevation Z is obtained from the DEM, and the coordinates of the original image are solved in reverse. Grayscale resampling: obtained through bilinear interpolation The pixel values ​​at the specified locations are used to fill in the orthophoto; the orthophoto is output: a single orthophoto without distortion is generated.

[0046] Design of algorithm for generating absolute coordinates of pile foundation center: Using models such as YOLOv5 to detect photovoltaic piles, the output bounding boxes are adapted to photovoltaic pile detection. Photovoltaic piles may vary in size and angle in the image. YOLOv5's adaptive anchor box calculation and adaptive image scaling can adapt to the detection of targets at different scales, effectively detecting photovoltaic piles.

[0047] Combining semantic segmentation to refine the profile of the photovoltaic (PV) pile improves positioning accuracy: In PV pile positioning, the bounding box output by target detection only provides a general location, while the actual profile of the PV pile may be irregular. Semantic segmentation can accurately separate the PV pile from the background, refining its profile and providing a foundation for subsequent precise positioning.

[0048] Improving positioning accuracy, semantic segmentation achieves pixel-level precision by classifying each pixel in an image. It captures detailed information about solar power plants, reducing positioning errors caused by blurred boundaries and complex backgrounds, thus significantly enhancing positioning accuracy.

[0049] Calculate the geometric center point based on the bounding box or segmentation results: Use semantic segmentation to obtain the set of pixel points of the pile outline, and determine the geometric center point by calculating the coordinates of these pixel points.

[0050] Convert image coordinates to absolute coordinates in the world coordinate system: based on the image pixel coordinate system. →Camera Coordinate System →UAV body coordinate system →Geographic coordinate system The four-layer coordinate system completes the calculation of the center coordinates of the pile foundation.

[0051] Regarding the overall software design (full-process automation + CAD engineering BIM overlay): Flight path generation and execution system: The management system automatically reads the coordinates of the CAD engineering pile foundations (including those already constructed), and based on the DJI SDK protocol, automatically generates the optimal flight path according to the requirements of image processing algorithms. The main algorithm for generating the optimal flight path is the Particle Swarm Optimization (PSO) algorithm. The on-site operator automatically acquires the flight path from the management system via the remote controller Pilot, and automatically uploads the pile foundation image data to the management system after collection.

[0052] Particle swarm optimization algorithm calculates the optimal trajectory: the objective function is optimized by the logic of "particle = candidate solution, position = parameter of solution, velocity = parameter to adjust direction". In UAV route optimization, the compatibility between the two must be clarified first.

[0053] Particle: One particle corresponds to one candidate UAV flight path, which consists of the coordinates (such as latitude, longitude, and altitude) of a series of consecutive waypoints, such as... ,in It is the first The three-dimensional coordinates of each waypoint.

[0054] The objective function includes: Objective 1: Shortest Path: Minimize the total distance of the flight path. Suitable for scenarios with limited range or requiring quick task completion. Example formula: , Where n is the number of waypoints, and the summation result is the total length of the route.

[0055] Objective 2: Minimize Energy Consumption: Combining the UAV energy consumption model, such as the relationship between energy consumption and flight speed, altitude difference, and payload, minimize the total energy consumption. Example formula: , The speed energy consumption coefficient, The energy consumption coefficient is the energy consumption coefficient based on altitude difference. Based on energy consumption, For the first The flight speed of the segment of the route.

[0056] Objective 3: Multi-objective optimization: If it is necessary to simultaneously satisfy "shortest path + low energy consumption + short time", multiple objectives can be transformed into a single objective through weighted summation, for example: , , , The weights are 1, and the sum is 1.

[0057] Set constraints: Ensure that the flight path is executable. UAV flight path optimization must meet physical and mission constraints; otherwise, the optimization results will be meaningless.

[0058] The constraints include: Endurance constraints: Total flight distance ≤ maximum range of the drone. Emergency power should be reserved, usually 80% of the maximum range.

[0059] Obstacle avoidance constraint: The distance from all waypoints in the flight path to obstacles (such as buildings, mountains, no-fly zones) is greater than or equal to the safety threshold.

[0060] Task constraints: If it is an inspection task, it must cover a designated area, and the overlap rate between the route and the area must be ≥95%; if it is a transportation task, it must accurately pass through the take-off and landing points and the target point.

[0061] Discrete waypoint characteristics of UAV routes: Standard PSO is suitable for continuous space optimization, while UAV routes consist of discrete waypoints.

[0062] Particle encoding: "Waypoint sequence encoding" is adopted, where each particle directly corresponds to a set of waypoint coordinates, rather than an abstract continuous variable, ensuring a one-to-one correspondence between particles and flight paths.

[0063] Speed ​​update constraint: When updating waypoint coordinates, the speed range must be limited, such as adjusting the x and y axes by no more than 100 meters per iteration step and the z axis by no more than 50 meters, to avoid waypoint jumps that would make the route unflyable.

[0064] Constraint handling: For particles that do not meet the constraints, a penalty term is added when calculating fitness. For example, if the fitness value for obstacle crossing is multiplied by 10, the algorithm is forced to eliminate invalid routes.

[0065] Iterative optimization process: From initialization to optimal output, initialization parameters are determined, the task area (e.g., boundary of offshore photovoltaic field), task points (e.g., pile foundation inspection points), and constraint parameters (maximum range, safe distance) are defined, and PSO parameters (number of particles) are set. Maximum number of iterations Inertial weight Learning factor ).

[0066] Generate initial particle swarm: Each particle randomly generates a set of waypoint sequences to ensure that the initial waypoints are within the mission area and initially meet the altitude and speed constraints.

[0067] Calculate fitness: For each particle, substitute it into the objective function to calculate the fitness value, and at the same time check the constraints. If the constraints are not met, add a penalty term.

[0068] Update individual best and global best: Each particle compares its current fitness with its own historical best (pbest), and if it is better, it updates pbest; all particles compare their current pbest with the global best (gbest), and if it is better, they update gbest.

[0069] Update speed and location: Update formula based on PSO speed:

[0070] Adjust the speed at each waypoint, then update the formula based on position:

[0071] Adjust waypoint coordinates while ensuring the new waypoints meet constraints.

[0072] Iteration Termination: If the maximum number of iterations is reached, or the fitness value of gbest shows no significant change (e.g., the change amount) after 10 consecutive iterations. If the condition is met, the iteration stops, and the route corresponding to gbest is output as the optimal path.

[0073] CAD Engineering BIM Overlay System: The management system automatically parses pile foundation images and obtains positioning information. By comparing them with CAD engineering drawings, it automatically marks the pile foundation construction error values ​​and displays them overlaid through BIM, making it convenient for construction personnel to adjust the pile foundation extension plate.

[0074] Data processing and accuracy assessment: Multi-line lidar point cloud data are fused and noise is removed.

[0075] Extract the coordinates of characteristic points of the pile foundation (such as the center of the pile top and the edge of the pile bottom) and calculate the coordinates of the center position of the pile foundation.

[0076] Accuracy assessment: Compare the measured coordinates with the design coordinates. Planar error should be controlled within ±5cm, and elevation error within ±3cm. If deviations exceed these limits, analyze the source of the error, adjust the measurement plan, and remeasure.

[0077] In some embodiments, the aforementioned high-precision positioning and measurement system for offshore photovoltaic power station pile foundations can also be implemented in the following ways.

[0078] In this embodiment, the high-precision positioning and measurement system for offshore photovoltaic power station pile foundations includes: Unmanned aerial vehicle platform: equipped with high-precision lidar, dual-frequency GNSSRTK module, and attitude sensor; Route planning module: Generates main routes and cross-validation routes based on the layout of the photovoltaic field and lidar parameters, supporting manual adjustment and automatic optimization; Point cloud preprocessing: filtering, registration; Pile foundation feature extraction: edge detection and geometric fitting algorithms are used; Coordinate correction: Enables coordinate transformation based on ground control points; Accuracy assessment module: Compares measured coordinates with design coordinates and generates an error report; Data storage and transmission: Measurement data is stored in real time on the drone's onboard SD card and simultaneously transmitted to the ground server via 4G / 5G network.

[0079] In some embodiments, the above-described high-precision positioning and measurement method for offshore photovoltaic power station pile foundations can also be implemented in the following ways.

[0080] In this embodiment, Figure 2 As shown, the high-precision positioning and measurement method for offshore photovoltaic power station pile foundations includes: Step 1: Measurement Preparation and Flight Data Acquisition. Import the coordinates of the engineering drawings of the constructed pile foundations into the UAV flight path planning system, and then... Figure 3 The particle swarm optimization algorithm shown generates the main flight path and the cross-validation flight path. Four ground control points are set up on the shore. Before executing the mission, the UAV scans the control points to complete the conversion and calibration between its own coordinate system and the engineering coordinate system.

[0081] Step 2: Locating the center point of the pile foundation and comparing and analyzing multi-source coordinates. The UAV system optimizes the calculation of the position of the center point of the pile foundation in the acquired images, and integrates the coordinates from manual on-site testing with the coordinates from the design drawings for comparative calculation and analysis.

[0082] In this example, as Figure 4 , Figure 5 As shown, the camera calibration algorithm can accurately obtain the lens focal length, aperture and lens design parameters. Using a checkerboard pattern as the calibration target, 30 calibration images at different angles are taken, and the camera intrinsic parameters and distortion coefficients are solved. Based on these parameters, image distortion correction is completed, providing lens-error-free image data for subsequent center point positioning.

[0083] In this example, as Figure 6 As shown, a digital elevation model (DEM) is generated from point cloud data collected by lidar using orthorectification technology, and an orthorectified image output grid is defined. The original image coordinates x and y are solved in reverse for each pixel X and Y of the orthorectified image using collinearity equations. Gray-scale resampling is completed using bilinear interpolation, and finally, a pile foundation orthorectified image with accurate geographic coordinates and no geometric distortion is generated.

[0084] In this example, as Figure 7As shown, based on orthophotos, the YOLOv5 target detection model is used to locate the pile foundation bounding box. The image pixel coordinates of the center point are calculated by the intersection of the diagonals of the bounding box. Then, through four coordinate system transformations, namely from the pixel coordinate system to the camera coordinate system (based on camera intrinsic parameters), from the camera coordinate system to the UAV body coordinate system (based on attitude sensor parameters), and from the UAV body coordinate system to the geographic coordinate system (based on GNSS coordinates), the absolute coordinates of the pile foundation center point are calculated using the UAV coordinate calculation algorithm. The positioning accuracy reaches ±10mm, meeting the accuracy requirements for offshore photovoltaic pile foundation measurement.

[0085] Step 3, as follows Figure 8 As shown, the management system automatically reads the pile foundation coordinates (including already constructed pile foundations) from the CAD engineering drawings and automatically generates flight paths using image processing algorithms. On-site operators automatically acquire the flight paths planned by the management system via remote control. After completing the acquisition of pile foundation image data, the system automatically uploads the data to the management system, achieving fully automated integration of the "flight path generation - acquisition - data acquisition - upload" process.

[0086] In this example, as Figure 9 As shown, the UAV map management system can automatically parse pile foundation image data and extract real-time pile foundation positioning information. It then compares this information with the pile foundation design coordinates in the CAD engineering drawings, automatically calculating and labeling the pile foundation construction error values. Furthermore, the error information is overlaid with the 3D model of the pile foundation through a BIM model, providing construction personnel with an intuitive and reliable reference for accurately adjusting the pile foundation extension plate.

[0087] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A high-precision positioning and measurement method for offshore photovoltaic power station pile foundations, characterized in that, Includes the following steps: S1: Based on the coordinates of the engineering drawings of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters; S2: Obtain the target image according to the main flight path and cross-validation flight path; perform camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center; S3: Obtain the pile foundation construction error value based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation.

2. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The objective function of the particle swarm optimization algorithm is: ， ， ， in, The objective function is... , , Let be the weights, and their sum is ; , and These represent the shortest path objective function, the low energy consumption objective function, and the shortest time objective function, respectively. , and Let n be the horizontal x-coordinate, horizontal y-coordinate, and altitude of the UAV; where n is the number of waypoints. The speed energy consumption coefficient, The energy consumption coefficient is the energy consumption coefficient based on altitude difference. Based on energy consumption, For the first The flight speed of the segment of the route.

3. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The constraints of the particle swarm optimization algorithm include endurance constraints, obstacle avoidance constraints, and task constraints; the endurance constraint is the total distance of the route. The maximum flight range of the drone is used as a reserve for emergency power; the obstacle avoidance constraint is the distance from all waypoints on the flight path to obstacles. Safety threshold; the task constraints are: if it is an inspection task, it covers a designated area, and the overlap rate between the route and the area is ≥95%; if it is a transportation task, it passes through the take-off and landing points and the target point.

4. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The velocity update formula for the particle swarm optimization algorithm is: ， The position update formula for the particle swarm optimization algorithm is: ， in, and Let represent the velocity of the i-th particle in the j-th dimension at generation t and generation (t+1); Indicates inertia weight; and Indicates the learning factor; and Represents a random number within the interval [0,1], enhancing the randomness of the algorithm's search; This represents the historical best position of the i-th particle in the j-th dimension; and This represents the position of the i-th particle in the j-th dimension at generation t and generation (t+1); This represents the globally optimal position in the j-th dimension of the entire particle swarm.

5. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The main flight path is used to make the UAV fly along the planned row direction of the photovoltaic array. The spacing between adjacent flight paths of the main flight path is equal to the effective scanning width of the UAV's lidar, which is used to ensure that the pile foundation is within the scanning range. The cross-verification flight path forms a 45° angle with the main flight path, and the flight altitudes of the main flight path and the cross-verification flight path are different. There are no fewer than two cross-verification flight paths, and all cross-verification flight paths intersect. The cross-verification flight path is used to eliminate changes in the UAV's flight attitude and lidar scanning angle errors by fusing data from multiple flight paths.

6. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The method for obtaining the coordinate transformation parameters is as follows: using RTK to obtain the known coordinates of multiple ground control points, and using the lidar of the UAV to obtain the lidar scan coordinates of multiple ground control points; using the known coordinates of the multiple ground control points and the lidar scan coordinates of the multiple ground control points to perform matching calculations to solve for the coordinate transformation parameters.

7. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The camera calibration process includes: Camera intrinsic parameters and distortion coefficients are obtained from multiple calibration images taken from different angles; The target image is corrected for image distortion using the camera intrinsic parameters and distortion coefficients to obtain a target image without lens error.

8. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The orthorectification process includes: A digital elevation model is obtained based on point cloud data; Based on the digital elevation model and the orthophoto output grid, the original image coordinates are solved in reverse for each pixel of the orthophoto using the collinearity equation. Based on the original image coordinates, grayscale resampling is performed using bilinear interpolation to obtain an orthophoto of the pile foundation with accurate geographic coordinates and no geometric distortion.

9. The high-precision positioning and measurement method for offshore photovoltaic power station pile foundations according to claim 1, characterized in that, The target detection process includes: Based on the orthophoto of the pile foundation, the YOLOv5 target detection model is used to locate the boundary box of the pile foundation. The image pixel coordinates of the center point are solved by using the intersection of the diagonals of the pile foundation boundary frame; The image pixel coordinates are converted into camera coordinates using camera intrinsic parameters; The camera coordinates are converted into UAV body coordinates using attitude sensor parameters; The coordinates of the UAV body are converted into geographic coordinates using the positioning system coordinates to obtain the absolute coordinates of the pile foundation center.

10. A high-precision positioning and measurement system for offshore photovoltaic power station pile foundations, characterized in that, The system includes: Flight path planning module: Based on the coordinates of the engineering drawings of the constructed pile foundations, the main flight path and cross-validation flight path are generated using the particle swarm optimization algorithm; the coordinates of the UAV are corrected using the known coordinates of multiple ground control points to obtain coordinate transformation parameters; Point cloud preprocessing, pile foundation feature and coordinate correction module: acquires target image according to the main flight path and cross-validation flight path; performs camera calibration, orthorectification and target detection on the target image according to the coordinate transformation parameters to obtain the absolute coordinates of the pile foundation center; Accuracy assessment module: Based on the absolute coordinates of the pile foundation center and the engineering drawing coordinates of the constructed pile foundation, the construction error value of the pile foundation is obtained.

Images