Unmanned aerial vehicle cooperative inspection laser deicing system and method

Abstract

The invention belongs to the technical field of wind power generation equipment maintenance, and particularly relates to an unmanned aerial vehicle cooperative inspection laser deicing system and method and an unmanned aerial vehicle cluster, the unmanned aerial vehicle cluster comprises a master control unmanned aerial vehicle and slave unmanned aerial vehicles, the master control unmanned aerial vehicle carries a binocular camera used for photographing and a laser radar used for distance measurement, and the slave unmanned aerial vehicle carries a laser radar used for distance measurement. The slave unmanned aerial vehicle is used for constructing a blade three-dimensional icing map, the slave unmanned aerial vehicle is provided with a pulse laser generator for heating an ice layer and a thermal infrared camera, the resolution ratio of the thermal infrared camera is 640 * 512 pixels, the temperature measurement precision is + / -0.5 DEG C, and the thermal infrared camera is used for collecting a blade surface temperature field in real time; the communication delay of the 5GMesh network is less than 100ms, and dynamic node access and real-time data synchronization are supported; the unmanned aerial vehicle carries the binocular camera and the laser radar to achieve three-dimensional precise modeling of the icing area, the positioning error is small, the detection efficiency is improved compared with manual detection, and the high-altitude operation risk brought to workers is also avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation equipment maintenance, and in particular relates to a laser deicing system and method for cooperative inspection by unmanned aerial vehicles. Background Art

[0002] In the field of wind power generation, wind turbine blades are core components for energy conversion. Ice formation on their surfaces is a major cause of reduced power generation efficiency and equipment failure. Statistics show that when ice thickness exceeds 2mm on blades, aerodynamic efficiency can drop by 15%-20%. In extreme cases, it can even lead to safety accidents such as blade breakage. Therefore, efficient and reliable blade de-icing technology is crucial for ensuring stable operation of wind turbine equipment in extremely cold and humid environments.

[0003] Problems with existing technologies: Traditional manual de-icing requires climbing, with a single maintenance cost exceeding 20,000 yuan and the risk of falling; mechanical de-icing equipment, such as helicopter sandblasting, can easily damage the blade coating, and the aerodynamic performance degrades after maintenance; existing drone liquid spray de-icing technology is greatly affected by environmental humidity, with efficiency dropping sharply below -10°C, and residual agents contaminating the blades; although the concept of laser de-icing has been proposed, existing solutions have defects such as insufficient energy control accuracy, high risk of thermal damage, large positioning errors, and lack of closed-loop verification, making it difficult to meet engineering needs. Summary of the Invention

[0004] The purpose of the present invention is to provide a UAV collaborative inspection laser deicing system and method, which can build an efficient, safe, and low-consumption unmanned deicing system, provide a breakthrough solution for the reliable operation of wind turbines in severe cold and high humidity environments, and have significant engineering application value and economic benefits.

[0005] The technical solutions adopted by the present invention are as follows: A UAV collaborative inspection laser deicing system, comprising: A drone swarm includes a master drone and slave drones. The master drone is equipped with a binocular camera for taking photos and a lidar for ranging, used to construct a three-dimensional ice map of the blades. The slave drones are equipped with a pulsed laser generator for heating the ice layer and a thermal infrared camera. The thermal infrared camera has a resolution of 640×512 pixels and a temperature measurement accuracy of ±0.5°C. The thermal infrared camera is used to collect the blade surface temperature field in real time. The 5GMesh network has a communication delay of less than 100ms and supports dynamic node access and real-time data synchronization. The central collaborative control de-icing module includes a laser de-icing module and a closed-loop verification module. The laser de-icing module uses pulse width modulation technology to control the output of the laser energy of the pulse laser generator. The laser de-icing module has a laser path planning algorithm based on the ice thickness cloud map, which is used to give priority to processing areas with an icing probability greater than 80%, including the leading edge and tip of the blade. The closed-loop verification module includes a temperature field real-time monitoring unit and an ice layer residue judgment unit. The temperature field real-time monitoring unit collects the blade surface temperature field in real time through the thermal infrared camera. The ice layer residue judgment unit determines whether the de-icing is completed based on the established "temperature gradient-ice layer residue" judgment model, and automatically triggers secondary fill light in the unqualified area.

[0006] The pulse laser generator has a wavelength of 1064 nm, a single pulse energy of less than 500 mJ, a power of less than 500 W, a laser beam divergence angle of less than 1.5 mrad, and a focused spot diameter of 5-10 mm.

[0007] In the "temperature gradient-residual ice layer" discrimination model, the threshold for determining that de-icing is complete is that the local temperature difference is less than 2°C.

[0008] A method for laser deicing using a drone for collaborative inspection includes the following steps: S1, deploys a drone cluster to the target wind farm, establishes a communication link through the 5GMesh network, and synchronizes the initial coordinate data; S2, the master drone scans the blade surface, constructs a 3D point cloud model, identifies the icing area, generates an ice thickness cloud map, and transmits it to the slave drones; S3, the laser path is planned from the UAV, the flight attitude is adjusted so that the laser beam is incident perpendicularly on the surface to be treated, and the pulse laser generator is started to perform the deicing operation; In step S4, the thermal infrared camera collects temperature field data and analyzes the de-icing effect based on the discriminant model. Secondary fill light is triggered in areas that do not meet the standard, and the step S3 operation is repeated until the thermal infrared image shows that the ice layer is completely cleared. S5: After completing de-icing in a single area, the drone cluster moves to the next icing area according to the preset path, repeating S2 to S4 until the inspection and de-icing of the entire blade is completed. The master drone summarizes the data, generates a report, and uploads it; In S6, ground personnel conduct remote assessments, adjust parameters or paths if necessary, and initiate secondary operations.

[0009] In S2, when constructing a three-dimensional point cloud model and identifying ice areas, the threshold is set to a reflectivity greater than 80% and a thickness greater than 2 mm, and the accuracy of the generated ice thickness cloud map is ±0.2 mm.

[0010] In the above S3, during the laser deicing operation, the laser temperature rise rate is 5±1°C / s, the laser beam incident angle deviation is less than 5°, and the scanning speed is 10-50 mm / s.

[0011] In S6, the de-icing effect is evaluated to have a residual ice layer of less than 0.5 mm and a blade temperature rise of less than 10°C.

[0012] The technical effects achieved by the present invention are: This invention, a collaborative innovation of human-machine cluster and laser de-icing technology, breaks through the safety and efficiency bottlenecks of traditional high-altitude manual maintenance and realizes the full process automation of wind turbine blade deicing.

[0013] The specific technical effects are reflected in the following: the drone is equipped with a binocular camera and a lidar to achieve accurate three-dimensional modeling of the ice area, with small positioning error. Compared with manual inspection, it improves the inspection efficiency and avoids the risk of high-altitude operations for workers.

[0014] The pulse laser provided in the present invention accurately controls energy output through pulse width modulation technology, so that the temperature rise rate is maintained at 5±1°C / s, effectively avoiding overheating of the blade base. The aerodynamic power loss after deicing is relatively small, the energy consumption is effectively reduced compared with traditional electric heating, and the energy utilization rate is also relatively high.

[0015] The thermal infrared camera provided in the present invention verifies the de-icing effect in real time, can identify residual ice layers less than 0.5 mm, and combines the "temperature gradient-ice layer residue" model to achieve high accuracy of secondary fill light, making the overall operation success rate relatively high.

[0016] The 5GMesh network set up in the present invention ensures the synchronization of multi-machine data and the accuracy of laser path planning and fill-light operation; the system is suitable for use in low-temperature and high-humidity environments, effectively reducing single maintenance costs, does not require chemical agents and is compatible with moderate rain conditions, and can cover a large range of icing wind fields.

[0017] In summary, the present invention constructs an efficient, safe, and low-consumption unmanned de-icing system, providing a breakthrough solution for the reliable operation of wind turbines in severe cold and high-humidity environments, and has significant engineering application value and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flowchart of the UAV cluster deicing operation of the present invention; Figure 2 This is a logic diagram for distinguishing ice residue in thermal infrared images according to the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is described in detail below with reference to the following examples. It should be understood that the following text is only used to describe one or more specific embodiments of the present invention and does not strictly limit the scope of protection of the present invention.

[0020] like Figure 1-Figure 2 As shown, a UAV collaborative inspection laser deicing system includes: A drone swarm consists of a master drone and slave drones. The master drone is equipped with a binocular camera for taking photos and a lidar for ranging, which is used to build a three-dimensional ice map of the blades. The slave drones are equipped with a pulsed laser generator for heating the ice layer and a thermal infrared camera. The thermal infrared camera has a resolution of 640×512 pixels and a temperature measurement accuracy of ±0.5°C. The thermal infrared camera is used to collect the blade surface temperature field in real time. The drone swarm is equipped with a microcomputer to verify the ice melting effect and determine whether laser treatment is needed. The pulsed laser generator has a wavelength of 1064nm, a single pulse energy of less than 500mJ, a power of less than 500W, a laser beam divergence angle of less than 1.5mrad, and a focused spot diameter of 5-10mm. The 5GMesh network has a communication delay of less than 100ms and supports dynamic node access and real-time data synchronization to supplement ice melting. The central collaborative control de-icing module includes a laser de-icing module and a closed-loop verification module. The laser de-icing module uses pulse width modulation technology to control the output of the pulsed laser generator laser energy. The laser de-icing module has a laser path planning algorithm based on the ice thickness cloud map, which is used to prioritize areas with an icing probability greater than 80%, including the leading edge and tip of the blade. The closed-loop verification module includes a real-time temperature field monitoring unit and an ice residue judgment unit. The real-time temperature field monitoring unit collects the blade surface temperature field in real time through a thermal infrared camera. The ice residue judgment unit is based on the established "temperature gradient-ice residue" judgment model. When the local temperature difference is less than 2°C, it is determined that de-icing is completed, and the non-compliant area automatically triggers secondary fill light.

[0021] The above structure breaks through the safety and efficiency bottlenecks of traditional high-altitude manual maintenance and realizes the automation of the entire process of wind turbine blade icing treatment. The drone is equipped with a binocular camera and a laser radar to achieve three-dimensional accurate modeling of the icing area, reducing positioning errors and improving detection efficiency compared to manual labor, effectively eliminating the risks of high-altitude operations. A method for laser deicing using a drone for collaborative inspection includes the following steps: S1 deploys a master drone equipped with a binocular camera and lidar and a cluster of slave drones equipped with a pulsed laser generator and thermal infrared camera to the target wind farm, establishes a communication link through the 5GMesh network, and synchronizes the initial coordinate data; In S2, the master drone flies along the preset inspection path at a distance of 10-50m from the blade surface. It scans the blade surface through the fusion of a binocular camera and a lidar, constructs a 3D point cloud model, and identifies ice areas. The threshold is set to a reflectivity greater than 80% and a thickness greater than 2mm. A cloud map of ice thickness with an accuracy of ±0.2mm is generated and synchronously transmitted to the slave drones.

[0022] S3: The laser path is planned from the UAV based on the ice thickness cloud map, with priority given to key areas such as the leading edge and tip of the blade. The flight attitude is adjusted so that the laser beam is incident perpendicularly on the surface to be treated, with an incident angle deviation of less than 5°. The pulsed laser generator is activated to locally heat the ice layer at a temperature rise rate of 5±1°C / s. The single pulse energy of the pulsed laser generator is less than 500mJ, the spot diameter is 5-10mm, and the scanning speed is 10-50mm / s. In step 4, a thermal infrared camera collects temperature data in the laser action area in real time and analyzes the de-icing effect based on a "temperature gradient-ice layer residue" discrimination model. If the de-icing effect does not meet the standard, a secondary fill-in light program is automatically triggered, and the de-icing operation in step 3 is repeated until the thermal infrared image shows that the ice layer has been completely cleared. S5: After completing de-icing in a single area, the drone cluster moves to the next icing area along a preset path and repeats steps S2 to S4 until the entire blade is inspected and de-iced. The master drone summarizes the de-icing area, energy consumption, and time consumed during the operation, generates a 3D visualization report, and uploads it to the ground control center. In S6, ground personnel remotely analyze the operation report and evaluate the de-icing effect. The goal of evaluating the de-icing effect is to ensure that the residual ice layer is less than 0.5 mm and the blade temperature rise is less than 10°C. If necessary, the laser parameters or inspection path are adjusted, and a secondary operation is initiated to ensure thorough de-icing.

[0023] Example 1 Scenario: In winter, the temperature remains around -30°C for a long time, and the ambient humidity is as high as 90%. The extreme low temperature and high humidity conditions cause wind turbine blades to freeze quickly, and the ice layer thickness grows extremely fast.

[0024] Operation process: After the master drone and the slave drones establish a communication link via the 5GMesh network, the master drone begins scanning the blades along the planned inspection path at a height of 15 meters from the blade surface. The binocular camera and lidar work stably, quickly constructing a 3D point cloud model, accurately identifying eligible ice areas, generating an ice thickness cloud map, and transmitting it to the slave drones. After receiving data from the drone, the single pulse energy of the pulse laser generator is set to 420mJ, the spot diameter is 7mm, the scanning speed is adjusted to 20mm / s, and the de-icing operation is started at a temperature rise rate of 4°C / s in response to the icing situation on the blades. During the de-icing process, the thermal infrared camera collects temperature field data in real time. Since heat dissipates quickly in low-temperature environments, the system uses the "temperature gradient-ice layer residue" discrimination model to quickly analyze the temperature data. When a local temperature difference greater than 2°C is detected, secondary fill light is immediately triggered. After multiple rounds of meticulous work, the ice layer on the entire blade was completely removed. After testing, the thickness of the remaining ice layer was only 0.3mm, and the temperature rise of the blade was controlled within 8°C, effectively completing the de-icing task in an extremely cold and humid environment.

[0025] Example 2 Scenario: The temperature fluctuates between -20°C and 0°C, and the humidity is maintained at 80%-95%. Ice on the wind turbine blades here is unevenly distributed and has wide variations in thickness.

[0026] Operation process: After the master drone and the slave drones establish a communication link via the 5GMesh network, the master drone begins scanning the blades along the planned inspection path at a height of 15 meters from the blade surface. The binocular camera and lidar work stably, quickly constructing a 3D point cloud model, accurately identifying eligible ice areas, generating an ice thickness cloud map, and transmitting it to the slave drones. Based on the ice thickness cloud map, the drone increased the single pulse energy to 480mJ and slowed the scanning speed to 15mm / s for thicker ice areas, such as the leading edge of the blades. For thinner ice areas, the single pulse energy was set to 380mJ and the scanning speed was increased to 25mm / s for de-icing. A thermal infrared camera monitored temperature changes throughout the process, accurately capturing temperature data in a high-humidity environment. Based on a "temperature gradient-ice residue" discrimination model, the de-icing effect was assessed in real time, and secondary lighting was promptly applied to areas that did not meet the standards. De-icing operations were successfully completed on all blades, with residual ice thickness less than 0.5mm and blade temperature rise below 10°C, effectively ensuring the normal operation of the wind turbine.

[0027] Example 3 Scenario: Winter is often hit by cold waves, with temperatures plummeting to -25°C and humidity reaching 100%. Wind turbine blades are not only severely frozen but also frosted.

[0028] Operation process: After the master drone and the slave drone establish a communication link through the 5GMesh network, the master drone begins scanning the blades along the planned inspection path at a height of 15m from the blade surface. The binocular camera and lidar work stably, quickly constructing a 3D point cloud model, accurately identifying the ice areas that meet the conditions, generating an ice thickness cloud map, and transmitting it to the slave drones. After the drone planned the laser path based on the cloud map, the de-icing process was initiated. Taking into account the impact of frost on de-icing, the system optimized the parameters of the pulsed laser generator, stabilizing the single pulse energy at 450mJ, the spot diameter at 8mm, and gradually melting the ice and frost at a temperature rise rate of 5°C / s. In a high-humidity and low-temperature environment, the thermal infrared camera overcomes water vapor interference and accurately collects temperature field data. The closed-loop verification module strictly controls the de-icing effect based on the "temperature gradient-ice residue" judgment model. Once the local temperature difference is found to be substandard, the secondary fill light is triggered immediately. After multiple rounds of meticulous de-icing and fill light operations, the ice and frost on the entire blade are completely removed. The thickness of the residual ice layer is less than 0.5mm, and the blade temperature rise is only 9°C. In the extreme low-temperature and high-humidity environment, it demonstrates powerful de-icing capabilities and stability.

[0029] The pulsed laser precisely controls energy output through PWM technology, maintaining a temperature rise rate of 5±1°C / s and a maximum blade surface temperature rise of only 6.8°C. After deicing, aerodynamic power loss is less than 1.5%, reducing energy consumption by over 90% compared to traditional electric heating. Single-pulse energy is less than 500mJ, and energy utilization reaches 85%. A thermal infrared camera verifies deicing effectiveness in real time, identifying residual ice layers less than 0.5mm. Combined with a "temperature gradient-ice residue" model, it achieves 100% accuracy in secondary fill-in illumination, resulting in an overall operation success rate of 98.7%. The 5GMesh network ensures multi-machine data synchronization delays of less than 100ms and positioning errors of less than 5cm, ensuring the accuracy of laser path planning and fill-in illumination operations. The system is suitable for low temperatures of -30°C to 0°C and humidity environments of 60% to 100%, reducing single-time maintenance costs by 75% compared to manual labor. It requires no chemicals and is compatible with moderate rain conditions, covering over 90% of the world's icing wind farms.

[0030] The foregoing is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained herein shall, unless otherwise specified or limited, be implemented in accordance with conventional means in the art.

Claims

1. A UAV collaborative inspection laser deicing system, characterized by: include: A drone cluster, comprising a master drone and slave drones. The master drone is equipped with a binocular camera for taking photos and a lidar for ranging. The slave drones are equipped with a pulsed laser generator for heating the ice layer and a thermal infrared camera for real-time acquisition of the blade surface temperature field. The central collaborative control de-icing module includes a laser de-icing module and a closed-loop verification module. The laser de-icing module uses pulse width modulation technology to control the output of laser energy of the pulse laser generator. The closed-loop verification module includes a temperature field real-time monitoring unit and an ice layer residual discrimination unit. The temperature field real-time monitoring unit uses the thermal infrared camera to collect the blade surface temperature field in real time. The ice layer residual discrimination unit is based on the established "temperature gradient-ice layer residual" discrimination model.

2. The UAV collaborative inspection laser deicing system according to claim 1 is characterized by: The pulse laser generator has a wavelength of 1064 nm, a single pulse energy of less than 500 mJ, a power of less than 500 W, a laser beam divergence angle of less than 1.5 mrad, and a focused spot diameter of 5-10 mm.

3. The UAV collaborative inspection laser deicing system according to claim 1 is characterized by: The drone cluster synchronizes data via the 5GMesh network.

4. The UAV collaborative inspection laser deicing system according to claim 1 is characterized by: The laser deicing module has a laser path planning algorithm based on an ice thickness cloud map.

5. The UAV collaborative inspection laser deicing system according to claim 1 is characterized by: The thermal infrared camera has a resolution of 640×512 pixels and a temperature measurement accuracy of ±0.5°C.

6. The UAV collaborative inspection laser deicing system according to claim 1 is characterized by: In the "temperature gradient-ice layer residue" discrimination model, the threshold for determining whether de-icing is complete is a local temperature difference of less than 2°C, and whether de-icing is complete or not is determined, and secondary fill light is automatically triggered in areas that do not meet the standard.

7. A method for UAV collaborative inspection laser deicing, based on the UAV collaborative inspection laser deicing system according to any one of claims 1 to 6, characterized in that: The following steps are involved: S1, deploys a drone cluster to the target wind farm, establishes a communication link through the 5GMesh network, and synchronizes the initial coordinate data; S2, the master drone scans the blade surface, constructs a 3D point cloud model, identifies the icing area, generates an ice thickness cloud map, and transmits it to the slave drones; S3, the laser path is planned from the UAV, the flight attitude is adjusted so that the laser beam is incident perpendicularly on the surface to be treated, and the pulse laser generator is started to perform the deicing operation; In step S4, the thermal infrared camera collects temperature field data and analyzes the de-icing effect based on the discriminant model. Secondary fill light is triggered in areas that do not meet the standard, and the step S3 operation is repeated until the thermal infrared image shows that the ice layer is completely cleared. S5: After completing de-icing in a single area, the drone cluster moves to the next icing area according to the preset path, repeating S2 to S4 until the inspection and de-icing of the entire blade is completed. The master drone summarizes the data, generates a report, and uploads it; In S6, ground personnel conduct remote assessments, adjust parameters or paths if necessary, and initiate secondary operations.

8. The method for laser deicing using a UAV in collaborative inspection according to claim 7, characterized in that: In S2, when constructing a three-dimensional point cloud model and identifying ice areas, the threshold is set to a reflectivity greater than 80% and a thickness greater than 2 mm, and the accuracy of the generated ice thickness cloud map is ±0.2 mm.

9. The method for laser deicing using a UAV in collaborative inspection according to claim 7, characterized in that: In the above S3, during the laser deicing operation, the laser temperature rise rate is 5±1°C / s, the laser beam incident angle deviation is less than 5°, and the scanning speed is 10-50 mm / s.

10. The method for laser deicing using a UAV in collaborative inspection according to claim 7, characterized in that: In S6, the de-icing effect is evaluated to have a residual ice layer of less than 0.5 mm and a blade temperature rise of less than 10°C.

Images