Fan blade coating path planning method, system and electronic device

Abstract

The application provides a fan blade coating path planning method, system and electronic equipment, the method comprises the following steps: a fan blade three-dimensional modeling step: obtaining three-dimensional point cloud data of the fan blade, and generating a three-dimensional model containing surface curvature distribution information based on the three-dimensional point cloud data; a region intelligent division step: based on the surface curvature distribution information of the three-dimensional model, using a clustering algorithm to adaptively divide the coating surface of the fan blade into a plurality of non-overlapping sub-regions, and generating corresponding coating operation parameters for each sub-region; a dynamic path planning step: for each sub-region and the coating operation parameters corresponding to each sub-region, generating a coating path of a UAV, and realizing coating of the fan blade based on the coating path. The coating efficiency is significantly improved, and the coating quality is accurately controllable.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, specifically to a method, system, and electronic device for wind turbine blade coating path planning. Background Technology

[0002] As a core component of wind power generation systems, the surface condition of wind turbine blades directly affects power generation efficiency and equipment lifespan. Large wind turbine blades can reach 60-80 meters in length and have a surface area exceeding 200 square meters. Long-term exposure to complex environments such as high-altitude strong winds, ultraviolet radiation, rain, snow, and salt spray makes them highly susceptible to surface corrosion, coating peeling, and micro-cracks. Industry statistics indicate that wind turbine blade maintenance costs account for 25%-35% of the total operation and maintenance costs of wind farms, with large-area coating operations, such as anti-corrosion coating replacement and repair material spraying, being one of the most significant maintenance scenarios. However, traditional coating methods and existing technologies have many significant drawbacks.

[0003] Currently, wind farms both domestically and internationally widely employ manual high-altitude coating methods. This involves technicians suspended from the wind turbine blades by baskets or ropes, using handheld spray guns for coating. This method suffers from extremely low efficiency, poor safety, and uneven coating quality. Coating a single wind turbine blade requires 3-5 workers working continuously for 2-3 days, resulting in a long full maintenance cycle. The risk of falls from heights is high, with an accident rate of approximately 0.5 incidents per 1,000 turbines per year. Operational stability is also difficult to guarantee in strong winds. Furthermore, manual operation is affected by physical strength and experience, leading to coating thickness errors of up to ±20%, with widespread instances of missed or over-coated areas. While the single-UAV coating technology, which has gradually developed in recent years, has replaced manual labor to some extent, it still faces bottlenecks such as limited operating range, flawed path planning, insufficient coating uniformity, and weak environmental adaptability. A single drone has a flight time of only 20-30 minutes, and coating a single wind turbine blade requires 8-12 hours. Existing fixed trajectories do not take into account the curvature changes of the wind turbine blade surface, and the coating omission rate in areas with abrupt curvature changes such as the blade tip is as high as 15%-20%. Fluctuations in the distance between the drone and the wind turbine blade surface will cause changes in the nozzle coverage radius, and traditional fixed step length paths are prone to material waste or omissions. High-altitude airflow disturbances will also cause the drone's attitude to shake, leading to an increase in coating thickness error.

[0004] To address the efficiency issues of single-drone systems, some studies have attempted multi-drone collaboration, but existing technologies are primarily geared towards inspection and logistics scenarios, failing to meet the specific needs of coating operations. These technologies suffer from several shortcomings: lack of surface adaptability, resulting in sub-region overlap exceeding 10%; lack of collaborative control, leading to trajectory conflict avoidance response times exceeding 200ms and a high risk of mid-air collisions; and the absence of a dynamic adjustment mechanism for coating parameters, making it impossible to adjust flight speed or nozzle parameters based on real-time coating results. With the rapid growth of global wind power capacity, exceeding 100GW of new installations in 2024, wind farms urgently require efficient, high-precision, and low-cost wind turbine blade coating technologies. The core technological challenges lie in how to achieve non-overlapping region division and task allocation for swarm drones based on the curved surface characteristics of wind turbine blades, how to dynamically adjust path step size and drone pose to ensure full coverage and uniformity of coating, and how to achieve real-time trajectory conflict avoidance and operation synchronization among drones through distributed collaborative control. Developing a drone swarm collaborative coating path planning system specifically for the complex curved surfaces of wind turbine blades is crucial to solving these problems. Summary of the Invention

[0005] This invention provides a method, system, and electronic device for wind turbine blade coating path planning, which can solve the problems of how to achieve non-overlapping area division and task allocation of swarm drones based on the curved surface features of wind turbine blades, how to dynamically adjust the path step size and drone pose to ensure full coverage and uniformity of coating, and how to achieve real-time trajectory conflict avoidance and operation synchronization among drones through distributed cooperative control.

[0006] The technical solution provided by this invention is as follows:

[0007] On the one hand, a method for planning the coating path of wind turbine blades is provided, the method comprising:

[0008] 3D modeling steps for wind turbine blades: Obtain 3D point cloud data of wind turbine blades, and reconstruct a 3D model containing surface curvature distribution information based on the 3D point cloud data;

[0009] The intelligent region division step is as follows: Based on the surface curvature distribution information of the three-dimensional model, a clustering algorithm is used to adaptively divide the coating surface of the wind turbine blade into multiple non-overlapping sub-regions, and corresponding coating operation parameters are generated for each sub-region.

[0010] Dynamic path planning steps: For each sub-region and the coating operation parameters corresponding to each sub-region, a coating path for the UAV is generated, and the coating of the wind turbine blades is realized based on the coating path.

[0011] In one alternative implementation, it further includes:

[0012] The real-time curvature change rate of the sub-region is obtained, and the step size of the coating path is dynamically adjusted based on the real-time curvature change rate of the sub-region.

[0013] The coating operation parameters and corresponding coating paths for the sub-regions are allocated to the UAV cluster, and distributed dynamic conflict avoidance operations are performed based on the real-time status information shared between UAVs through a communication network; and

[0014] During the coating process, the coating thickness is monitored in real time, and the coating parameters are dynamically adjusted based on the monitoring results.

[0015] In one optional implementation, the intelligent region partitioning step involves using a clustering algorithm to adaptively divide the coated surface of the wind turbine blade into multiple non-overlapping sub-regions. This includes employing a K-means clustering algorithm, where the distance metric function of the K-means clustering algorithm integrates spatial coordinates and curvature values, and iteratively calculates the sub-region area balance and curvature similarity as optimization objectives until the K-means clustering algorithm converges, thereby outputting the final sub-region partitioning result.

[0016] In one optional implementation, the intelligent region partitioning step further includes:

[0017] After the division is completed, the task priority is set based on the average curvature value of each sub-region;

[0018] Sub-regions with an average curvature value higher than a preset threshold are given a higher job priority and are given priority in coating tasks.

[0019] In one optional implementation, the dynamic path planning step involves dynamically adjusting the step size of the path using the following formula:

[0020]

[0021] Wherein, S is the step size, R is the spray radius of the coating device, α is the preset overlap rate threshold, and θ is the angle between the UAV and the normal to the surface of the wind turbine blade at the path point; the step size S decreases as the angle θ increases, so as to achieve coating path densification in high curvature areas.

[0022] In one optional implementation, the dynamic path planning step generates a coating path for the UAV for each sub-region and the corresponding coating operation parameters, including using different path modes for sub-regions with different curvature characteristics:

[0023] For sub-regions with curvature change rate within a first range, a raster-style path is used, which consists of a series of parallel straight-line paths to improve coverage efficiency.

[0024] For sub-regions with a curvature change rate within a second range, a spiral path is adopted, which unfolds in a spiral line starting from the center or edge of the sub-region to better fit the complex curved surface contour; wherein, the second range is larger than the first range.

[0025] In one optional implementation, based on the real-time shared status information between UAVs via a communication network, a distributed dynamic conflict avoidance operation is performed, including:

[0026] The system monitors ambient wind speed in real time using an onboard wind speed sensor.

[0027] When the wind speed exceeds the first threshold, control all drones to hover in their current position and maintain their pose;

[0028] When the wind speed exceeds a second threshold, all drones are controlled to terminate their current mission and return to the ground base station along a preset safe path, wherein the second threshold is greater than the first threshold; and

[0029] Each drone shares its location information in real time through a communication network;

[0030] When the distance between any two drones is less than a preset safety threshold, calculate the repulsive force from other drones and the attractive force from the target point for each drone.

[0031] The resultant force of the repulsive force and the attractive force is decomposed into tangential and normal velocity adjustment components, and the flight control system corrects the flight speed vector and direction of the UAV in real time based on the velocity adjustment components to maintain a safe distance.

[0032] In an optional implementation, a coating quality closed-loop control step is further included, comprising:

[0033] The coating thickness is collected by an infrared thickness sensor at a preset frequency to obtain the deviation between coating thicknesses;

[0034] When a deviation exceeding -5% of the target thickness is detected, the drone is controlled to reduce its flight speed or the discharge flow rate of the coating device is increased.

[0035] When a deviation exceeding +5% of the target thickness is detected, the drone is controlled to increase its flight speed or reduce the output flow of the coating device.

[0036] In another aspect, a wind turbine blade coating path planning system is provided to implement the wind turbine blade coating path planning method as described in any one of the above claims, the system comprising:

[0037] The task planning layer includes a 3D modeling module and a region partitioning module. The 3D modeling module is used to perform the 3D modeling steps of the wind turbine blades, and the region partitioning module is used to perform the intelligent region partitioning steps.

[0038] The cluster coordination control layer includes a path planning module, which is used to perform dynamic path planning steps.

[0039] On the other hand, an electronic device is provided, comprising:

[0040] Memory, used to store computer programs;

[0041] A processor is configured to execute a computer program stored in the memory, wherein when the computer program is executed, it implements any of the wind turbine blade coating path planning methods described above.

[0042] The method provided in this embodiment of the invention has at least the following beneficial effects:

[0043] This invention, through "curvature-based adaptive partitioning," ensures precise matching between each subtask and the geometric features of the wind turbine blade from the task source, solving the problem of incompatibility caused by applying planar planning methods to curved surfaces in existing technologies. By "dynamically adjusting the path step size," it compensates for surface changes in real time during the coating process, directly addressing the characteristic of wind turbine blades having "large tip curvature and small root curvature," fundamentally eliminating the geometric root causes of missed coating and material waste. This method elevates wind turbine blade coating from a manual task relying on experience to a quantifiable, optimizable, and automated engineering process. Attached Figure Description

[0044] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0045] Figure 1 This is a flowchart illustrating a wind turbine blade coating path planning method provided in an embodiment of the present invention.

[0046] Figure 2 This is a schematic diagram of the system framework for a wind turbine blade coating path planning method provided in an embodiment of the present invention.

[0047] Figure 3 This is a flowchart illustrating a specific embodiment of a wind turbine blade coating path planning method provided by the present invention. Detailed Implementation

[0048] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0049] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this invention is for describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0050] Please see Figure 1 On one hand, embodiments of the present invention provide a method for wind turbine blade coating path planning, the method comprising:

[0051] Step S1 of 3D modeling of wind turbine blades: Obtain 3D point cloud data of wind turbine blades, and reconstruct a 3D model containing surface curvature distribution information based on the 3D point cloud data.

[0052] Region intelligent partitioning step S2: Based on the surface curvature distribution information of the 3D model, a clustering algorithm is used to adaptively partition the coating surface of the wind turbine blade into multiple non-overlapping sub-regions, and corresponding coating operation parameters are generated for each sub-region.

[0053] Dynamic path planning step S3: For each sub-region and the corresponding coating operation parameters, generate the coating path of the UAV, and implement the coating of the wind turbine blades based on the coating path.

[0054] This invention, through "curvature-based adaptive partitioning," ensures precise matching between each subtask and the geometric features of the wind turbine blade from the task source, solving the problem of incompatibility caused by applying planar planning methods to curved surfaces in existing technologies. By "dynamically adjusting the path step size," it compensates for surface changes in real time during the coating process, directly addressing the characteristic of wind turbine blades having "large tip curvature and small root curvature," fundamentally eliminating the geometric root causes of missed coating and material waste. This method elevates wind turbine blade coating from a manual task relying on experience to a quantifiable, optimizable, and automated engineering process.

[0055] Furthermore, the embodiments of the present invention, through the above-described method, have at least the following advantages: 1. Significantly improved coating efficiency: By using a drone swarm collaborative operation mode, the traditional operation time of a single drone or manual labor is greatly shortened. Taking a 200-square-meter wind turbine blade as an example, the coating time for a single wind turbine blade can be shortened from 8-12 hours for a single drone or 3-5 days for manual labor to less than 3.5 hours, improving efficiency by more than 50%. Simultaneously, it supports parallel operation of multiple wind turbine blades, meeting the large-scale maintenance needs of large wind farms and effectively reducing power generation losses caused by wind farm downtime for maintenance. 2. Precise and controllable coating quality: Based on the dynamic path planning and real-time monitoring feedback mechanism of the three-dimensional model of the wind turbine blade, the entire process of coating quality control is achieved. Dynamic step size adjustment ensures that the path coverage of areas with varying curvature (such as the blade tip and leading edge) is seamless, reducing the missed coating rate to 0; the deviation between coating thicknesses can be controlled within ±5% (traditional manual error ±20%), significantly improving material uniformity. Real-time monitoring and parameter compensation mechanisms using infrared sensors avoid local over-coating or under-coating, ensuring the consistency of the coating's protective performance. 3. **Safe and Reliable Cluster Collaboration:** Employing distributed communication and an improved artificial potential field method for conflict avoidance, drones achieve millisecond-level location sharing (communication latency ≤100ms) via 5G / LoRa networking, with a safe distance control accuracy of ±0.3m, completely eliminating the risk of collisions during cluster operations. An environmental adaptation unit monitors wind speed, temperature, and other parameters in real time, automatically hovering or returning to base in extreme weather conditions, improving operational safety by over 90% compared to manual high-altitude operations. 4. **Significantly Reduced Operating Costs:** The fully autonomous operation mode reduces human intervention by over 90%, avoiding the risks and labor costs of high-altitude operations (the manual maintenance cost for a single wind turbine blade is approximately 20,000 RMB, while the cost of cluster drone operations can be reduced to 5,000 RMB). Dynamic path planning and material flow control technologies increase material utilization by 30%, further reducing consumable costs. The modular system design supports compatibility with multiple drone brands, avoiding redundant equipment investment. 5. **Strong Adaptability to Complex Scenarios:** For the complex hyperboloid features of wind turbine blades, 3D modeling and intelligent region segmentation technology achieve precise matching between sub-region tasks and drone performance. Whether in the low-curvature area at the blade root or the high-curvature area at the blade tip, the coating effect can be ensured through adaptive step size and posture adjustment. It is compatible with wind turbine blades of different specifications from 60 to 80 meters and supports various coating material types such as anti-corrosion and repair, making it suitable for a wide range of applications. 6. Outstanding intelligence and scalability: The task planning layer, which integrates AI algorithms, can automatically optimize area division and path strategy, reducing manual debugging time; the ground monitoring platform provides full-process visualization and manual intervention interfaces, supporting real-time monitoring and data traceability of multi-machine collaborative operation status. The system reserves sensor interfaces, which can be expanded to integrate coating adhesion detection, defect identification and other functions, and can be upgraded to an integrated intelligent maintenance system of "detection-coating-acceptance" in the future.

[0056] In one alternative implementation, the real-time curvature change rate of the sub-region is obtained, and the step size of the coating path is dynamically adjusted based on the real-time curvature change rate of the sub-region.

[0057] The coating operation parameters and corresponding coating paths for sub-regions are allocated to the UAV cluster, and distributed dynamic conflict avoidance operations are performed based on the real-time status information shared between UAVs through the communication network; and

[0058] During the coating process, the coating thickness is monitored in real time, and the coating parameters are dynamically adjusted based on the monitoring results.

[0059] In one optional implementation, the region intelligent partitioning step involves using a clustering algorithm to adaptively partition the wind turbine blade coating surface into multiple non-overlapping sub-regions. This includes employing a K-means clustering algorithm, where the distance metric function of the K-means clustering algorithm integrates spatial coordinates and curvature values, and iteratively calculates the sub-region area balance and curvature similarity as optimization objectives until the K-means clustering algorithm converges, thereby outputting the final sub-region partitioning result.

[0060] The "integrated spatial coordinates and curvature values" method in this invention ensures that the divided sub-regions are not only physically continuous, but more importantly, have highly consistent internal surface characteristics. This provides a precise prerequisite for "one-click configuration" of the most suitable coating operation parameters (such as path mode and reference step size) for each region, avoiding the arbitrariness of manual division. Meanwhile, the "area balance" optimization directly serves load balancing, preventing a single UAV mission area from becoming a bottleneck in the overall system's operation time, thereby ensuring improved cluster parallel efficiency.

[0061] In one optional implementation, the region intelligent partitioning step further includes:

[0062] After the division is completed, the task priority is set based on the average curvature value of each sub-region;

[0063] Sub-regions with an average curvature value higher than a preset threshold are given a higher job priority and are given priority in coating tasks.

[0064] This invention addresses the specific condition where high-curvature areas (such as blade tips and leading edges) on wind turbine blades are more prone to corrosion and more sensitive to coating quality. It directly links the geometric characteristic of "curvature" with the engineering requirement of "maintenance risk." By prioritizing operations in high-risk areas, it ensures that the most critical and vulnerable areas are protected in unforeseen circumstances such as drone battery depletion or material shortages. This achieves intelligent allocation of maintenance resources and improves the overall reliability of the wind farm operation.

[0065] In one optional implementation, the step size of the coating path is dynamically adjusted in the dynamic path planning step using the following formula:

[0066]

[0067] Where S is the step size, R is the spray radius of the coating device, α is the preset overlap rate threshold, and θ is the angle between the UAV and the normal to the surface of the wind turbine blade at the path point; the step size S decreases as the angle θ increases in order to achieve path densification in high curvature areas.

[0068] The physical significance of this formula lies in the fact that when a drone flies over a high-curvature region of a wind turbine blade (such as the blade tip), its attitude will inevitably tilt, causing the angle θ between the nozzle and the normal of the wind turbine blade to increase. At this time, the formula automatically reduces the step size S through the cos(θ) term, making the path points denser, thereby compensating for the insufficient coverage that may be caused by tilted projection. This is equivalent to giving the drone the ability to understand surface geometry and fine-tune its operation in real time, accurately solving the core contradiction that fixed-step path planning cannot balance efficiency in flat areas and quality in complex areas on a specific workpiece like a wind turbine blade.

[0069] In one optional implementation, the dynamic path planning step generates a coating path for the UAV for each sub-region and the corresponding coating operation parameters, including adopting different path modes for sub-regions with different curvature characteristics:

[0070] For sub-regions with curvature change rate within the first range, a raster path is used, which consists of a series of parallel straight-line paths to improve coverage efficiency.

[0071] For sub-regions with curvature change rate within the second range, a spiral path is adopted, which starts from the center or edge of the sub-region and unfolds in a spiral line to better fit the complex surface contour; wherein, the second range is larger than the first range, for example, the curvature change rate in the first range is 10% and the curvature change rate in the second range is 15%.

[0072] For the flat area in the middle of the wind turbine blade, a grid path with simple calculations and short travel distance is adopted to maximize efficiency; for areas with complex curvature changes such as the blade tip and root, a spiral path that can naturally conform to the complex contours and avoid sharp turns is adopted to maximize coverage quality. This "sub-pattern" planning is not a simple superposition of known paths, but a targeted technical selection based on a deep understanding of the curved surface characteristics of the wind turbine blade, and its effect is an optimal balance between efficiency and quality at the global level.

[0073] In one optional implementation, based on the real-time shared status information between UAVs via a communication network, a distributed dynamic conflict avoidance operation is performed, including:

[0074] The system monitors ambient wind speed in real time using an onboard wind speed sensor.

[0075] When the wind speed exceeds the first threshold, control all drones to hover in their current position and maintain their pose;

[0076] When the wind speed exceeds the second threshold, control all drones to terminate the current mission and return to the ground base station along a preset safe path. The second threshold is greater than the first threshold.

[0077] In this embodiment, a two-level response mechanism ensures safety: when the wind speed is strong but still manageable (e.g., the first threshold of 8 m / s), the system hovers and waits for the wind conditions to improve, avoiding forced operation that could lead to uneven coating or even impact on the wind turbine blades; when the wind speed exceeds the safety limit (e.g., the second threshold of 15 m / s), the system automatically returns to base to protect the expensive drone equipment. This is equivalent to installing an "environmental awareness brain" on the entire system, greatly improving the robustness and economy of operation in real, variable wind farm environments. For example, hovering is precisely positioned by the flight control system through data fusion of GPS (Global Positioning System) and IMU (Inertial Measurement Unit); returning to base involves executing a pre-set "return point" navigation procedure in the flight control system. Hovering maintains the current position and altitude through the flight control system; returning to base uses a pre-stored sequence of route coordinates for navigation, which enhances the robustness and safety of the wind turbine blade coating path planning system in real-world environments.

[0078] In one optional implementation, the distributed dynamic conflict avoidance operation, based on the real-time shared status information between UAVs via a communication network, further includes:

[0079] Each drone shares its location information in real time through a communication network;

[0080] When the distance between any two drones is less than a preset safety threshold, calculate the repulsive force from other drones and the attractive force from the target point for each drone.

[0081] The resultant force of repulsive and attractive forces is decomposed into tangential and normal velocity adjustment components. The flight control system then uses these velocity adjustment components to correct the UAV's flight speed vector and direction in real time, thereby maintaining a safe distance.

[0082] It should be noted that traditional obstacle avoidance methods may directly track the direction of the resultant force, leading to path oscillations. This method decomposes the resultant force: the tangential component is used for speed adjustment (accelerating to quickly leave the danger zone or decelerating to wait), and the normal component is used for smooth steering. This decomposed control makes the UAV's obstacle avoidance maneuvers smoother and more stable in narrow spaces such as the sides of wind turbine blades, greatly reducing the interference to coating uniformity during violent UAV movements and significantly reducing the risk of collisions due to overshoot in dense swarms. This is achieved by the flight control system receiving the tangential and normal components and mapping them to control commands for the power motor and servo motors, respectively.

[0083] Furthermore, the method may also include a step coating quality closed-loop control step and a cluster collaborative control step to achieve dynamic conflict avoidance, etc.

[0084] The specific steps of the coating quality closed-loop control include: acquiring the coating thickness at a preset frequency using an infrared thickness sensor to obtain the deviation between coating thicknesses;

[0085] When a deviation exceeding -5% of the target thickness is detected, the drone is controlled to reduce its flight speed or the discharge flow rate of the coating device is increased.

[0086] When a deviation exceeding +5% of the target thickness is detected, the drone is controlled to increase its flight speed or reduce the output flow of the coating device.

[0087] This embodiment can correct unavoidable disturbances in actual operation, such as changes in drone speed due to high-altitude wind speed fluctuations or uneven material output due to changes in material viscosity. For example, when the thickness is detected to be too thin, the wind turbine blade coating path planning system will simultaneously reduce the flight speed and increase the flow rate to deposit more material per unit area. This real-time feedback and compensation transforms the coating quality control from an "open-loop" system that relies on operator experience to a "closed-loop" automatic control based on physical measurements. This is the core technological guarantee for ensuring uniform and consistent coating across wind turbine blades that are tens of meters long.

[0088] Please see Figure 2 Furthermore, a wind turbine blade coating path planning system is provided to implement any of the above-mentioned wind turbine blade coating path planning methods. The system includes:

[0089] The task planning layer includes a 3D modeling module and a region partitioning module. The 3D modeling module is used to perform the 3D modeling steps of the wind turbine blades, and the region partitioning module is used to perform the intelligent region partitioning steps.

[0090] This invention discloses a path planning system for large-area coating of wind turbine blades using a swarm of unmanned aerial vehicles (UAVs) in collaborative operations. It achieves efficient collaborative coating of multiple UAVs through a three-layer architecture of "task planning - swarm collaboration - coating execution." The specific technical solution is as follows:

[0091] Task Planning Layer: Wind Turbine Blade Modeling and Intelligent Region Allocation

[0092] The task planning layer is the system's "decision-making center," responsible for decomposing complex coating tasks into executable sub-tasks. First, the wind turbine blades are scanned in 3D using LiDAR or binocular vision sensors to acquire point cloud data. Then, the 3D modeling module uses a Poisson surface reconstruction algorithm to construct a 3D model of the wind turbine blade with an accuracy of ±2mm. The 3D model includes key features such as the blade root, tip, leading edge, and trailing edge, as well as the surface curvature distribution. The region division module uses an improved K-means clustering algorithm, with "curvature similarity" and "area balance" as optimization objectives, to divide the wind turbine blade surface into N non-overlapping sub-regions (N = number of drones, dynamically adjusted according to the wind turbine blade surface area, e.g., 4 sub-regions for a 200m² wind turbine blade). During the division process, special areas such as bolt holes and damaged areas on the wind turbine blade are automatically avoided, and each sub-region is assigned a priority (e.g., the tip corrosion risk is high, so the priority is set to level 1). The mission planning layer includes a mission parameter configuration module that synchronously sets coating process parameters, including material thickness (e.g., 80-120μm), UAV flight altitude (0.8-1.2m relative to the wind turbine blade surface), nozzle flow rate (5-10ml / s), and path overlap threshold (30%-40%), to ensure that the sub-region mission matches the UAV performance.

[0093] The cluster collaborative control layer includes a path planning module and a conflict avoidance module. The path planning module is used to execute dynamic path planning steps, and the conflict avoidance module is used to execute cluster collaborative control steps.

[0094] Furthermore, the cluster collaborative control layer serves as the system's "collaborative brain," enabling trajectory coordination and conflict avoidance among UAVs. The global path planning module generates a baseline path for each sub-region: a spiral trajectory (for regions with gentle curvature) or a grid-like trajectory (for complex curved surfaces) is used from the blade root to the blade tip. The path's starting point is uniformly set at the center of the blade root, and the ending point is at the edge of the blade tip, ensuring comprehensive coverage. The dynamic step-size calculation module adjusts the path interval based on the real-time curvature of the wind turbine blade surface, achieving adaptive optimization through the following formula.

[0095]

[0096] Where S is the dynamic step size, R is the spray radius of the coating device, α is the preset overlap rate threshold, and θ is the angle between the UAV and the normal to the wind turbine blade surface at the path point. The conflict avoidance module adopts an improved artificial potential field method, that is, the UAVs share their location, speed and task progress in real time through 5G or LoRa networking (communication latency ≤100ms). When the distance between two UAVs is less than the safety threshold (2m), the conflict avoidance module automatically calculates the repulsive force and the target attractive force, and adjusts the trajectory through the velocity vector decomposition method (such as the UAV shifting 0.5m towards the sub-region boundary) to ensure that there is no collision in the swarm operation.

[0097] The coating execution layer includes multiple drones equipped with coating devices and monitoring sensors, which perform coating operations and provide data feedback to achieve closed-loop control of coating quality.

[0098] For example, the coating execution layer is the system's "execution terminal," where a drone equipped with a coating device and multiple sensors completes the operation according to a planned path. The attitude adjustment unit, through a six-axis gimbal and an IMU (Inertial Measurement Unit), detects the angle between the drone and the normal to the wind turbine blade surface in real time. When the deviation exceeds ±3°, it drives the servo motor to adjust the nozzle attitude (response time ≤50ms) to ensure vertical material spraying. The coating monitoring unit integrates an infrared thickness sensor (sampling frequency 10Hz) to collect coating thickness data in real time and obtain deviations between coating thicknesses. If a deviation exceeds ±5% (e.g., target 80μm, measured 72μm), it immediately feeds back to the collaborative control layer, achieving dynamic compensation by adjusting flight speed (0.5-2m / s) or nozzle flow rate (±1ml / s). The environmental adaptation unit is equipped with a wind speed sensor. When a wind speed exceeding 8m / s is detected, an emergency hovering mechanism is triggered (hovering accuracy ±0.3m), resuming operation after the wind speed decreases; in extreme weather conditions, it automatically returns to a preset landing point to ensure equipment safety.

[0099] The ground monitoring platform communicates with the mission planning layer and the cluster collaborative control layer to display the UAV's pose, coating progress, and coating thickness distribution heat map in a visual manner in real time, and to receive intervention commands issued by users through a graphical interface.

[0100] The system in this embodiment of the invention is equipped with a ground monitoring platform, which receives cluster operation data via industrial Ethernet and displays the drone's pose, coating progress (e.g., area A is 75% complete), coating thickness distribution heat map, and abnormal alarms (e.g., "drone thickness deviation exceeds limit"). The ground monitoring platform supports manual intervention. When the automatically planned path needs optimization, the boundaries of sub-areas or the starting point of the path can be adjusted by dragging with the mouse; in emergencies, all drones can be paused with a single click to ensure that the operation remains controllable.

[0101] In another aspect, an electronic device is provided, comprising:

[0102] Memory, used to store computer programs;

[0103] The processor is used to execute a computer program stored in memory, and when the computer program is executed, it implements any of the above-mentioned wind turbine blade coating path planning methods.

[0104] On another front, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements any of the aforementioned wind turbine blade coating path planning methods. Detailed implementation method:

[0106] The following details the implementation steps of a wind turbine blade coating path planning system, using a practical application scenario as an example:

[0107] 1. 3D modeling of wind turbine blades and configuration of task parameters.

[0108] Please see Figure 3 Step 1.1: Use a drone equipped with a lidar to perform a surround scan of the wind turbine blades while the wind turbine is stopped (flight radius 50m, height level with the wind turbine blades) to obtain point cloud data (point density ≥100 points / cm²).

[0109] Step 1.2: Remove noise points using software, generate a 3D model of the wind turbine blade using the Poisson reconstruction algorithm, and export it as an STL format (S Tereo Lithography, a file format used for stereolithography computer-aided design software; STL format has both text and binary code forms). The 3D model includes the blade root, blade tip, and surface curvature changes.

[0110] Step 1.3: Set the coating operation parameters in the task planning layer: material is superhydrophobic coating (thickness 100μm), drone swarm size is 4, overlap rate threshold is 35%, safety distance is 2m, and operating temperature range is -10℃~40℃.

[0111] 2. Sub-region division and task allocation.

[0112] Step 2.1: Import the 3D model of the wind turbine blade into the region division module, set the clustering parameters (number of clusters 4, number of iterations 50), and automatically divide the surface of the wind turbine blade (200m²) into 4 sub-regions: Region A (blade root section, 55m², curvature 0.02-0.05 / m), Region B (left wing middle section, 48m², curvature 0.05-0.08 / m), Region C (right wing middle section, 47m², curvature 0.05-0.08 / m), and Region D (blade tip section, 50m², curvature 0.08-0.15 / m).

[0113] Step 2.2: Assign sub-regions to 4 drones using an algorithm, with priority D region > B region = C region > A region, to ensure priority operation in high curvature areas.

[0114] 3. Cluster collaborative coating operation process.

[0115] Step 3.1: The drone cluster takes off from the ground base station and flies to the starting point of the wind turbine blade area according to the planned path (the starting point of drone 1 in area A is the center of the blade root (0,0), and the starting point of drone 4 in area D is 1m from the blade tip (15,0)).

[0116] Step 3.2: The global path planning module generates a grid-like path (step size 0.5m) for UAV 1 and a spiral path (initial step size 0.3m) for UAV 4. The planned paths are sent to the UAV flight controller through the ROS system (Robot Operating System).

[0117] Step 3.3: During the operation, the dynamic step size module calculates the step size in real time at the point of maximum tip curvature in zone D (coordinates (28, 0.5)). At this time, R = 0.6m, α = 35%, θ = 5°, and the calculated value is S = 0.6 × (1 0.35)×cos(5°)≈0.38m, the step size is automatically adjusted to 0.38m.

[0118] Step 3.4: When the distance between UAV 2 and UAV 3 at the boundary of the sub-region (coordinates (10,3)) is close to 1.8m, the conflict avoidance module triggers obstacle avoidance: UAV 2 shifts 0.6m to the inside of area B, UAV 3 decelerates by 0.2m / s, the trajectory adjustment time is 0.4s, and the distance after obstacle avoidance is restored to 2.5m.

[0119] Step 3.5: The coating monitoring unit detects that the thickness at a certain point in area D is 75μm (target 100μm), and feeds it back to the collaborative control layer. The flight speed of UAV 4 decreases from 1.5m / s to 1.2m / s, the nozzle flow rate increases from 8ml / s to 9ml / s, and the thickness recovers to 98μm after 5 seconds.

[0120] 4. Job completion and data archiving.

[0121] Step 4.1: After all sub-areas are coated, the UAV returns along the original path, and the ground monitoring platform generates a coating quality report, including the total operation time (3.5 hours per wind turbine blade), average thickness (98μm, error ±4%), missed coating rate (0%), and material consumption (12L).

[0122] Step 4.2: Data is automatically uploaded to the wind farm operation and maintenance system, supporting historical data traceability and process optimization.

[0123] Through the above implementation methods, this system achieves high efficiency, precision, and safety in large-area coating of wind turbine blades, which is significantly better than traditional manual and single-drone operations, providing technical support for the maintenance of new energy equipment.

[0124] This invention presents an intelligent task partitioning method for wind turbine blades based on the curved surface features of the blades. By improving the K-means clustering algorithm, and using the curvature distribution and area balance of the three-dimensional model of the wind turbine blades as a basis, the surface of the wind turbine blades is adaptively divided into non-overlapping sub-regions, enabling dynamic allocation of UAV swarm tasks. During the partitioning process, special areas (such as bolt holes and damaged areas) are automatically avoided and priorities are set to ensure that high-risk areas (such as blade tips) are prioritized for operation, thus solving the problem of overlapping or missing sub-regions caused by traditional planar partitioning.

[0125] A dynamic path step size adjustment mechanism based on surface curvature coupling is adopted, namely, a step size calculation formula based on nozzle spray radius, overlap rate threshold, and the angle between the UAV and wind turbine blade normals is proposed: S=R (1 α) cos(θ) dynamically adjusts the path interval in real time according to the curvature of the wind turbine blade surface (such as shortening the step length in the high curvature area of ​​the blade tip to 60%-80% of the baseline value), and adds path density compensation for curvature abrupt changes in areas such as the leading edge and trailing edge to ensure full coverage and no missed coating of complex curved surfaces.

[0126] An improved artificial potential field method is adopted to achieve real-time location sharing among drones through a 5G / LoRa distributed communication network (latency ≤100ms). The repulsive force and target attraction of drones in the cluster are calculated, and the trajectory is dynamically adjusted through velocity vector decomposition to ensure the minimum safe distance and solve the collision risk caused by the lag in centralized scheduling.

[0127] It integrates a six-axis gimbal attitude adjustment unit and an infrared thickness monitoring unit to detect the angle between the surface normal of the drone and the wind turbine blade and the coating thickness in real time. When the thickness deviation exceeds ±5%, it automatically adjusts the flight speed or nozzle flow rate to form a closed-loop control of "monitoring-feedback-compensation" to ensure coating uniformity.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for planning the coating path of wind turbine blades, characterized in that, The method includes: 3D modeling steps for wind turbine blades: Obtain 3D point cloud data of wind turbine blades, and reconstruct a 3D model containing surface curvature distribution information based on the 3D point cloud data; The intelligent region division step is as follows: Based on the surface curvature distribution information of the three-dimensional model, a clustering algorithm is used to adaptively divide the coating surface of the wind turbine blade into multiple non-overlapping sub-regions, and corresponding coating operation parameters are generated for each sub-region. Dynamic path planning steps: For each sub-region and the coating operation parameters corresponding to each sub-region, generate a coating path for the UAV, and perform coating on the wind turbine blades based on the coating path; Also includes: The real-time curvature change rate of the sub-region is obtained, and the step size of the coating path is dynamically adjusted based on the real-time curvature change rate of the sub-region. The coating operation parameters and corresponding coating paths for the sub-regions are allocated to the UAV cluster, and distributed dynamic conflict avoidance operations are performed based on the real-time status information shared between UAVs through a communication network; and During the coating process, the coating thickness is monitored in real time, and the coating parameters are dynamically adjusted based on the monitoring results. The step size of the coating path is dynamically adjusted using the following formula: Where S is the step size, R is the spray radius of the coating device, α is the preset overlap rate threshold, and θ is the angle between the UAV and the normal to the surface of the wind turbine blade at the path point; the step size S decreases as the angle θ increases, so as to achieve coating path densification in high curvature areas. In the dynamic path planning step, a coating path for the UAV is generated for each sub-region and the corresponding coating operation parameters, including: For sub-regions with curvature change rate within a first range, a raster path is used, which consists of a series of parallel straight line paths; For sub-regions with a rate of change of curvature within a second range, a spiral path is adopted, which unfolds in a spiral shape starting from the center or edge of the sub-region; wherein the second range is larger than the first range; Based on real-time status information shared between UAVs via a communication network, distributed dynamic conflict avoidance operations are performed, including: The system monitors ambient wind speed in real time using an onboard wind speed sensor. When the wind speed exceeds the first threshold, control all drones to hover in their current position and maintain their pose; When the wind speed exceeds a second threshold, all drones are controlled to terminate their current mission and return to the ground base station along a preset safe path, wherein the second threshold is greater than the first threshold; and Each drone shares its location information in real time through a communication network; When the distance between any two drones is less than a preset safety threshold, calculate the repulsive force from other drones and the attractive force from the target point for each drone. The resultant force of the repulsive force and the attractive force is decomposed into tangential and normal velocity adjustment components, and the flight control system corrects the flight speed vector and direction of the UAV in real time based on the velocity adjustment components.

2. The wind turbine blade coating path planning method according to claim 1, characterized in that, In the intelligent region partitioning step, a clustering algorithm is used to adaptively divide the coated surface of the wind turbine blade into multiple non-overlapping sub-regions. This includes: using the K-means clustering algorithm, where the distance metric function of the K-means clustering algorithm integrates spatial coordinates and curvature values, and iteratively calculates the sub-region area balance and curvature similarity as optimization objectives until the K-means clustering algorithm converges, thereby outputting the final sub-region partitioning result.

3. The wind turbine blade coating path planning method according to claim 2, characterized in that, The intelligent region partitioning step also includes: After the division is completed, the task priority is set based on the average curvature value of each sub-region; Sub-regions with an average curvature value higher than a preset threshold are given a higher job priority and are given priority in coating tasks.

4. The wind turbine blade coating path planning method according to claim 1, characterized in that, It also includes a closed-loop control step for coating quality: The coating thickness is collected by an infrared thickness sensor at a preset frequency to obtain the deviation between coating thicknesses; When the deviation is detected to exceed -5% of the target thickness, the drone is controlled to reduce its flight speed or the discharge flow rate of the coating device is increased. When the deviation is detected to exceed +5% of the target thickness, the drone is controlled to increase its flight speed or reduce the output flow of the coating device.

5. A wind turbine blade coating path planning system, characterized in that, The system for implementing the wind turbine blade coating path planning method as described in any one of claims 1-4 includes: The task planning layer includes a 3D modeling module and a region partitioning module. The 3D modeling module is used to perform the 3D modeling steps of the wind turbine blades, and the region partitioning module is used to perform the intelligent region partitioning steps. The cluster coordination control layer includes a path planning module, which is used to perform dynamic path planning steps.

6. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute a computer program stored in the memory, wherein when the computer program is executed, it implements the wind turbine blade coating path planning method according to any one of claims 1-4.

Images