Wind power blade damage detection device

Abstract

The invention discloses a wind power blade damage detection device, and relates to the technical field of wind power blade detection, and the device comprises a detection module which is responsible for carrying out the comprehensive detection of a wind power blade, and a solar power supply module which is responsible for collecting energy through a solar panel, and storing the energy in a battery, so as to improve the cruising ability of an unmanned plane body. The data fusion and analysis module is responsible for integrating the data of the detection module to form a comprehensive wind power blade damage detection report, and the real-time monitoring and feedback module is responsible for transmitting the detected data to a monitoring center in real time. Through the ultrasonic detection unit, the visual detection unit and the infrared detection unit, wind power blade damage detection in multiple modes can be realized, so that the defects of a single detection method are overcome, different modal information extracted by multiple detection methods is fully utilized, the detection precision is improved, accurate identification of the surface damage and the internal damage of the wind power blade is realized, and the detection efficiency is improved. And accurate three-dimensional positioning of the damage is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine blade detection, and in particular to a wind turbine blade damage detection device. Background Art

[0002] As the global demand for renewable energy continues to grow, wind power generation, as a clean and renewable form of energy, has developed rapidly around the world. As a key component of wind turbines, the performance of wind turbine blades directly affects the efficiency and safety of wind power generation. Wind turbine blades are in complex and changeable environments for a long time, such as high wind speed, extreme temperature, humidity, salt spray, etc. These factors will cause varying degrees of damage to the blades. Common types of blade damage include fatigue damage, collision damage, corrosion damage, etc. These damages will not only affect the efficiency of wind power generation, but may even threaten the safe operation of wind turbines in severe cases. The use of drones to detect wind turbine blade damage mainly uses drones as flight platforms, equipped with high-definition cameras, infrared thermal imagers, lidars and other sensor equipment to conduct comprehensive and accurate digital monitoring of wind turbine blades. Through its flight capabilities, drones can observe the surface of blades at close range and from multiple angles, capturing subtle damage that is difficult to detect with the naked eye.

[0003] In the existing technology, it is difficult to integrate the three differentiated detection technologies of ultrasonic detection, visual detection and infrared detection to achieve high-precision identification and precise spatial positioning of wind turbine blade structural damage. This detection technology effectively overcomes the limitations of a single detection method in recognition accuracy and positioning accuracy, and significantly enhances the comprehensive efficiency of damage detection. However, the additional hardware equipment introduced by the above-mentioned detection technology directly increases the load of the drone, thereby increasing the energy consumption level. At the same time, the three detection methods run in parallel, further aggravating power consumption. The application of deep learning algorithms, as the core of the detection algorithm, is computationally intensive and also leads to a significant increase in energy consumption. It is difficult to ensure that the solar panels are aligned with the sun with the maximum effective area, that is, it is difficult to improve the photoelectric conversion efficiency, which may affect the wind turbine blade damage detection. Therefore, it is necessary to propose a wind turbine blade damage detection device. Summary of the invention

[0004] The purpose of the present invention is to solve the problem of achieving high-precision identification and precise spatial positioning of wind turbine blade structural damage by integrating three differentiated detection technical means. This detection technical means effectively overcomes the limitations of a single detection method in identification accuracy and positioning accuracy, and significantly enhances the comprehensive effectiveness of damage detection. However, the additional hardware equipment introduced by the above-mentioned detection technical means directly increases the load of the UAV, thereby increasing the energy consumption level. At the same time, the three detection methods run in parallel, further aggravating power consumption. The application of deep learning algorithms, as the core of the detection algorithm, is computationally intensive and also leads to a significant increase in energy consumption. It is difficult to ensure that the solar panels are aligned with the sun with the maximum effective area, that is, it is difficult to improve the photoelectric conversion efficiency, which may affect the detection of wind turbine blade damage. A wind turbine blade damage detection device is proposed.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A wind turbine blade damage detection device, comprising:

[0007] Detection module: responsible for comprehensive inspection of wind turbine blades;

[0008] Solar power module: responsible for collecting energy through solar panels and storing it in batteries to improve the endurance of the drone.

[0009] Data fusion and analysis module: responsible for integrating the detection module data to form a comprehensive wind turbine blade damage detection report;

[0010] Real-time monitoring and feedback module: responsible for transmitting the detected data to the monitoring center in real time;

[0011] A drone body, wherein the solar power supply module is arranged on the upper part of the drone body, a first angle control motor is fixedly installed on the upper part of the drone body, a first rotating shaft is arranged at the output end of the first angle control motor, a fixing sleeve is fixedly connected to the side of the first rotating shaft away from the first angle control motor, a second angle control motor is fixedly installed on the upper part of the drone body, a worm is arranged at the output end of the second angle control motor, a rotating sleeve is fixedly connected to the side of the fixing sleeve close to the second angle control motor, the rotating sleeve and the worm are rotationally connected, a turbine is meshedly connected to one side of the worm, the turbine is fixedly connected to the second rotating shaft, the second rotating shaft and the fixing sleeve are rotationally connected, the second rotating shaft is fixedly connected to the second connecting plate, a solar panel for powering the drone body is fixedly connected to the upper part of the second connecting plate, a high-definition camera is fixedly connected to the upper part of the drone body, an infrared sensor is fixedly connected to the drone body, and an ultrasonic sensor is fixedly connected to the drone body;

[0012] Among them, by adjusting the solar panels, it is ensured that the solar panels are aimed at the sun with the largest effective area, thereby maximizing the photoelectric conversion efficiency.

[0013] The above technical solution further includes:

[0014] The first rotating shaft is rotationally connected to the drone body, the worm is rotationally connected to the drone body, and the worm is rotationally connected to the first rotating shaft.

[0015] A first connecting plate is fixedly connected to the upper part of the drone body, a third telescopic plate is fixedly connected to a side of the first connecting plate close to the high-definition camera, the third telescopic plate is arranged on the upper part of the high-definition camera, the third telescopic plate is slidably connected to the second telescopic plate, and the second telescopic plate is slidably connected to the first telescopic plate.

[0016] A cylinder is fixedly mounted on the upper portion of the third telescopic plate, and a fixing block is provided at the output end of the cylinder, and the fixing block is fixedly connected to the first telescopic plate.

[0017] A cleaning control motor is fixedly mounted on the upper portion of the first connecting plate, a threaded rod is provided at the output end of the cleaning control motor, the threaded rod is rotationally connected to the drone body, and the cleaning plate is threadedly connected to the threaded rod.

[0018] A plurality of cleaning brushes for cleaning the surface of the high-definition camera are fixedly connected to one side of the cleaning plate close to the high-definition camera. The cleaning plate is slidably connected to a fixing rod. The upper part of the fixing rod is fixedly connected to the first connecting plate, and the lower part of the fixing rod is fixedly connected to the drone body.

[0019] The detection module includes an ultrasonic detection unit, a visual detection unit and an infrared detection unit. The ultrasonic detection unit transmits and receives ultrasonic waves through multiple ultrasonic sensors to detect whether there are cracks or structural defects in the wind turbine blades. The visual detection unit captures images or videos of wind turbine blade surface damage through a high-definition camera. The infrared detection unit detects the temperature distribution on the surface of the wind turbine blade through an infrared sensor to perform damage detection.

[0020] Among them, by integrating three differentiated detection technologies: ultrasonic detection, visual detection and infrared detection, high-precision identification and precise spatial positioning of wind turbine blade structural damage can be achieved.

[0021] The ultrasonic detection unit realizes three-dimensional positioning of damage through phased array ultrasonic detection, transmits and receives ultrasonic waves simultaneously through multiple ultrasonic probes, and processes and analyzes the information to obtain information of different depths and positions inside the blade, thereby realizing three-dimensional imaging and damage positioning.

[0022] The data fusion and analysis module uses the YOLO model to fuse the information collected by different detection units. The specific steps of the YOLO model fusing the information collected by different detection units are:

[0023] Data preprocessing: Collect the raw data of wind turbine blades from different detection units, remove noise, outliers and duplicate data from the raw data, standardize the data of different detection units to the same size, and then mark them, that is, select the annotation box containing the damage and mark the damage type;

[0024] YOLO model training: The preprocessed ultrasonic, visual and infrared image data are synthesized into a data set, which is then split into training set, validation set and test set in proportion. The training set is used to train the constructed YOLO model, and the validation set data is used to perform parameter optimization operations to find the optimal YOLO model. Finally, the performance of the optimal YOLO model trained is judged based on the test set data to determine the final YOLO model;

[0025] Model application and deployment: Apply the trained damage detection model to the actual wind turbine blade damage detection task. By inputting new ultrasonic, visual and infrared data, the model can automatically identify the damaged area and type of the blade.

[0026] Among them, the three data sets cannot have the same image data to prevent the selection of a false optimal model.

[0027] The present invention has the following beneficial effects:

[0028] 1. In the present invention, when the drone body is used to inspect the wind turbine blades, the cooperation of the first rotating shaft and the worm gear can ultimately realize the all-round angle adjustment of the solar panel, so that the solar panel can always face the sun to better receive sunlight, thereby maximizing the power generation efficiency, facilitating the improvement of the endurance of the drone body, and facilitating the drone body to carry out detailed inspection of the wind turbine blades in all directions and at multiple angles to ensure that nothing is missed.

[0029] 2. In the present invention, multiple methods of wind turbine blade damage detection can be realized through ultrasonic detection units, visual detection units and infrared detection units, so as to overcome the shortcomings of a single detection method, make full use of different modal information extracted by multiple detection methods, improve detection accuracy, and realize accurate identification of surface damage and internal damage of wind turbine blades, as well as accurate three-dimensional positioning of damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a schematic diagram of the structure of a wind turbine blade damage detection device proposed by the present invention;

[0031] Figure 2It is a schematic diagram of the first structure in the present invention;

[0032] Figure 3 It is a second structural schematic diagram of the present invention;

[0033] Figure 4 for Figure 1 A schematic diagram of the structure enlargement in the middle;

[0034] Figure 5 for Figure 3 A magnified schematic diagram of the structure at B in the middle;

[0035] Figure 6 It is a system block diagram of the present invention.

[0036] In the figure: 1. UAV body; 2. First angle control motor; 3. First rotating shaft; 4. Fixed sleeve; 5. Second angle control motor; 6. Solar panel; 7. Turbine; 8. Worm; 9. Infrared sensor; 10. Ultrasonic sensor; 11. High-definition camera; 12. First connecting plate; 13. First telescopic plate; 14. Second telescopic plate; 15. Third telescopic plate; 16. Cylinder; 17. Fixed block; 18. Fixed rod; 19. Cleaning control motor; 20. Second rotating shaft; 21. Second connecting plate; 22. Rotating sleeve; 23. Threaded rod; 24. Cleaning plate; 25. Cleaning brush. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] like Figure 1 - Figure 6 As shown, a wind turbine blade damage detection device comprises:

[0039] Detection module: responsible for comprehensive inspection of wind turbine blades;

[0040] Solar power supply module: responsible for collecting energy through the solar panel 6 and storing it in the battery to improve the endurance of the drone body 1;

[0041] Data fusion and analysis module: responsible for integrating the detection module data to form a comprehensive wind turbine blade damage detection report;

[0042] Real-time monitoring and feedback module: responsible for transmitting the detected data to the monitoring center in real time;

[0043] The drone body 1 and the solar power supply module are arranged on the upper part of the drone body 1. A first angle control motor 2 is fixedly installed on the upper part of the drone body 1. A first rotating shaft 3 is arranged at the output end of the first angle control motor 2. A fixing sleeve 4 is fixedly connected to the side of the first rotating shaft 3 away from the first angle control motor 2. A second angle control motor 5 is fixedly installed on the upper part of the drone body 1. A worm 8 is arranged at the output end of the second angle control motor 5. A rotating sleeve 22 is fixedly connected to the side of the fixing sleeve 4 close to the second angle control motor 5. The rotating sleeve 22 is rotationally connected to the worm 8. A turbine 7 is meshedly connected to one side of the worm 8. The turbine 7 is fixedly connected to the second rotating shaft 20. The second rotating shaft 20 is rotationally connected to the fixing sleeve 4. The second rotating shaft 20 is fixedly connected to a second connecting plate 21. A solar panel 6 for powering the drone body 1 is fixedly connected to the upper part of the second connecting plate 21. A high-definition camera 11 is fixedly connected to the upper part of the drone body 1. An infrared sensor 9 is fixedly connected to the drone body 1. An ultrasonic sensor 10 is fixedly connected to the drone body 1.

[0044] The working principle of a wind turbine blade damage detection device proposed by the present invention is that the real-time monitoring and feedback module transmits the detected data to the ground station or monitoring center in real time for timely processing and feedback, and when damage or abnormality is detected, the alarm system will notify the operator or system to take further action. When the angle of the solar panel 6 needs to be adjusted, the first angle control motor 2 is started, and the first angle control motor 2 drives the first rotating shaft 3 set at the output end to rotate, and the first rotating shaft 3 drives the fixedly connected fixed sleeve 4 to rotate. The fixed sleeve 4 is fixedly connected with a rotating sleeve 22 on one side close to the second angle control motor 5, and the rotating sleeve 22 and the first rotating shaft 3 are both rotatably connected to the worm 8;

[0045] That is, the rotation of the first rotating shaft 3 will drive the fixing sleeve 4 to tilt toward the side close to the infrared sensor 9 or away from the infrared sensor 9, and the rotation of the fixing sleeve 4 will drive the second rotating shaft 20 connected in rotation and the second connecting plate 21 fixedly connected to the second rotating shaft 20 to tilt, that is, the second connecting plate 21 will drive the solar panel 6 fixedly connected on the upper part to tilt, that is, the tilt angle of the solar panel 6 can be adjusted, and when the second angle control motor 5 is started, the second angle control motor 5 will drive the worm 8 set at the output end to rotate, and the rotation of the worm 8 will cause the meshing turbine 7 to rotate, and the rotation of the turbine 7 will drive the fixedly connected second rotating shaft 20 to rotate;

[0046] The rotation of the second rotating shaft 20 will cause the fixedly connected second connecting plate 21 to tilt toward a position close to the high-definition camera 11 or away from the high-definition camera 11, and will drive the solar panel 6 to tilt toward a position close to the high-definition camera 11 or away from the high-definition camera 11, so that the angle of the solar panel 6 can be fully adjusted to provide power for the drone body 1, thereby ensuring the endurance of the drone body 1, and at the same time providing power for the detection algorithm and the three detection methods to run in parallel. Through the cooperation of the infrared sensor 9, the ultrasonic sensor 10, and the high-definition camera 11, multiple detection methods can be used simultaneously to achieve high-precision recognition and precise spatial positioning of wind turbine blade structural damage, effectively overcoming the limitations of a single detection method in recognition accuracy and positioning accuracy, and significantly enhancing the comprehensive effectiveness of damage detection.

[0047] In one embodiment, for the above-mentioned first rotating shaft 3, the first rotating shaft 3 and the drone body 1 are rotationally connected, the worm 8 and the drone body 1 are rotationally connected, and the worm 8 and the first rotating shaft 3 are rotationally connected.

[0048] In one embodiment, for the above-mentioned drone body 1, a first connecting plate 12 is fixedly connected to the upper part of the drone body 1, a third telescopic plate 15 is fixedly connected to the side of the first connecting plate 12 close to the high-definition camera 11, the third telescopic plate 15 is arranged on the upper part of the high-definition camera 11, the third telescopic plate 15 is slidably connected to the second telescopic plate 14, and the second telescopic plate 14 is slidably connected to the first telescopic plate 13.

[0049] In one embodiment, for the third telescopic plate 15 , a cylinder 16 is fixedly mounted on the upper portion of the third telescopic plate 15 , a fixing block 17 is disposed at the output end of the cylinder 16 , and the fixing block 17 is fixedly connected to the first telescopic plate 13 .

[0050] The working principle of a wind turbine blade damage detection device proposed in the present invention is that when the cylinder 16 is started, the cylinder 16 will cause the fixed block 17 set at the output end to drive the first telescopic plate 13 to slide, and when the first telescopic plate 13 slides to the inside of the second telescopic plate 14, it will drive the second telescopic plate 14 to slide to the inside of the third telescopic plate 15 at the same time, that is, the lengths of the first telescopic plate 13, the second telescopic plate 14 and the third telescopic plate 15 can be adjusted, and part of the direct sunlight can be blocked to make the light softer, thereby preventing the high-definition camera 11 from overexposure when collecting images and videos and causing loss of details.

[0051] In one embodiment, for the above-mentioned first connecting plate 12, a cleaning control motor 19 is fixedly installed on the upper part of the first connecting plate 12, and a threaded rod 23 is provided at the output end of the cleaning control motor 19. The threaded rod 23 is rotationally connected to the drone body 1, and the threaded rod 23 is threadedly connected to a cleaning plate 24.

[0052] In one embodiment, for the above-mentioned cleaning plate 24, a plurality of cleaning brushes 25 for cleaning the surface of the high-definition camera 11 are fixedly connected to the side of the cleaning plate 24 close to the high-definition camera 11, and the cleaning plate 24 is slidably connected to a fixing rod 18, and the upper part of the fixing rod 18 is fixedly connected to the first connecting plate 12, and the lower part of the fixing rod 18 is fixedly connected to the drone body 1.

[0053] The working principle of a wind turbine blade damage detection device proposed in the present invention is to start the cleaning control motor 19, which will cause the threaded rod 23 set at the output end to rotate. The rotation of the threaded rod 23 will cause the threaded cleaning plate 24 to move up and down. When the cleaning plate 24 moves, it will drive multiple cleaning brushes 25 fixedly connected to one side of the high-definition camera 11 to move. When the cleaning brushes 25 move, they will have a cleaning effect on the surface of the high-definition camera 11, thereby ensuring the clarity of the high-definition camera 11 when collecting images and videos. The fixed rod 18 to which the cleaning plate 24 is slidably connected will keep the cleaning plate 24 stable when it moves.

[0054] In one embodiment, for the above-mentioned detection module, the detection module includes an ultrasonic detection unit, a visual detection unit and an infrared detection unit. The ultrasonic detection unit transmits and receives ultrasonic waves through multiple ultrasonic sensors 10 to detect whether there are cracks or structural defects in the wind turbine blade. The visual detection unit captures images or videos of surface damage of the wind turbine blade through a high-definition camera 11. The infrared detection unit detects the temperature distribution on the surface of the wind turbine blade through an infrared sensor 9 to perform damage detection.

[0055] In one embodiment, for the above-mentioned ultrasonic detection unit, the ultrasonic detection unit realizes three-dimensional positioning of damage through phased array ultrasonic detection, transmits and receives ultrasonic waves simultaneously through multiple ultrasonic probes, and processes and analyzes this information to obtain information at different depths and positions inside the blade, thereby realizing three-dimensional imaging and damage positioning.

[0056] The working principle of a wind turbine blade damage detection device proposed in the present invention is that the visual detection unit captures images or videos of the surface of the wind turbine blade through a high-definition camera 11, and processes and analyzes the acquired image data in real time, and finally detects and identifies damaged areas or abnormalities on the surface of the wind turbine blade; sends ultrasonic waves to the surface of the wind turbine blade through an ultrasonic sensor 10, processes and analyzes the received ultrasonic signals, and finally detects cracks or structural defects inside the blade; detects the temperature distribution on the surface of the wind turbine blade through an infrared sensor 9. When there is damage inside the wind turbine blade, the temperature distribution on the surface of the blade will change due to uneven heat conduction, and abnormal hot spots or temperature changes will be found, and finally the internal damage of the blade will be identified. The different modal information extracted by multiple detection methods is fully utilized to improve the detection accuracy, realize accurate identification of surface damage and internal damage of the wind turbine blade, and realize three-dimensional imaging and damage positioning through phased array ultrasonic detection.

[0057] In one embodiment, for the above-mentioned data fusion and analysis module, the data fusion and analysis module uses the YOLO model to fuse the information collected by different detection units. The specific steps of the YOLO model fusing the information collected by different detection units are:

[0058] Data preprocessing: Collect the raw data of wind turbine blades from different detection units, remove noise, outliers and duplicate data from the raw data, standardize the data of different detection units to the same size, and then mark them, that is, select the annotation box containing the damage and mark the damage type;

[0059] YOLO model training: The preprocessed ultrasonic, visual and infrared image data are synthesized into a data set, which is then split into training set, validation set and test set in proportion. The training set is used to train the constructed YOLO model, and the validation set data is used to perform parameter optimization operations to find the optimal YOLO model. Finally, the performance of the optimal YOLO model trained is judged based on the test set data to determine the final YOLO model;

[0060] Model application and deployment: The trained damage detection model is applied to the actual wind turbine blade damage detection task. By inputting new ultrasonic, visual and infrared data, the model can automatically identify the damaged area and type of the blade.

[0061] The working principle of a wind turbine blade damage detection device proposed in the present invention is that data preprocessing screens out images with higher quality and better clarity from the collected images, removes blurred, repeated or irrelevant images, and performs enhancement processing on the screened images, such as contrast adjustment and sharpening, so as to improve the visual effect of the image and the accuracy of feature extraction. During the training process of the YOLO model, the detection accuracy and generalization ability of the model can be improved by adjusting model parameters, optimizing algorithms, etc. When evaluating the YOLO model, it is necessary to check whether the performance indicators of the model, such as detection accuracy and recall rate, meet the requirements. When the deployed YOLO model is used to perform damage detection on the wind turbine blade image, the model will automatically identify the damaged area in the image and mark the corresponding category and confidence.

[0062] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A wind turbine blade damage detection device, characterized in that: include: Detection module: responsible for comprehensive inspection of wind turbine blades; Solar power supply module: responsible for collecting energy through solar panels (6) and storing it in batteries to improve the endurance of the drone body (1); Data fusion and analysis module: responsible for integrating the detection module data to form a comprehensive wind turbine blade damage detection report; Real-time monitoring and feedback module: responsible for transmitting the detected data to the monitoring center in real time; A drone body (1), wherein the solar power supply module is arranged on the upper part of the drone body (1), a first angle control motor (2) is fixedly mounted on the upper part of the drone body (1), a first rotating shaft (3) is arranged at the output end of the first angle control motor (2), a fixed sleeve (4) is fixedly connected to the side of the first rotating shaft (3) away from the first angle control motor (2), a second angle control motor (5) is fixedly mounted on the upper part of the drone body (1), a worm (8) is arranged at the output end of the second angle control motor (5), a rotating sleeve (22) is fixedly connected to the side of the fixed sleeve (4) close to the second angle control motor (5), and the rotating sleeve (22) is connected to the rotating sleeve (22). The worm gears (8) are rotatably connected to each other, one side of the worm gear (8) is meshedly connected to a turbine (7), the turbine (7) is fixedly connected to a second rotating shaft (20), the second rotating shaft (20) and the fixed sleeve (4) are rotatably connected, the second rotating shaft (20) is fixedly connected to a second connecting plate (21), the upper part of the second connecting plate (21) is fixedly connected to a solar panel (6) for supplying power to the drone body (1), the upper part of the drone body (1) is fixedly connected to a high-definition camera (11), the drone body (1) is fixedly connected to an infrared sensor (9), and the drone body (1) is fixedly connected to an ultrasonic sensor (10).

2. A wind turbine blade damage detection device according to claim 1, characterized in that: The first rotating shaft (3) is rotationally connected to the drone body (1), the worm (8) is rotationally connected to the drone body (1), and the worm (8) is rotationally connected to the first rotating shaft (3).

3. A wind turbine blade damage detection device according to claim 1, characterized in that: A first connecting plate (12) is fixedly connected to the upper part of the drone body (1); a third telescopic plate (15) is fixedly connected to a side of the first connecting plate (12) close to the high-definition camera (11); the third telescopic plate (15) is arranged on the upper part of the high-definition camera (11); the third telescopic plate (15) is slidably connected to a second telescopic plate (14); and the second telescopic plate (14) is slidably connected to the first telescopic plate (13).

4. A wind turbine blade damage detection device according to claim 3, characterized in that: A cylinder (16) is fixedly mounted on the upper part of the third telescopic plate (15), and a fixing block (17) is provided at the output end of the cylinder (16), and the fixing block (17) is fixedly connected to the first telescopic plate (13).

5. A wind turbine blade damage detection device according to claim 3, characterized in that: A cleaning control motor (19) is fixedly mounted on the upper portion of the first connecting plate (12); a threaded rod (23) is provided at the output end of the cleaning control motor (19); the threaded rod (23) is rotatably connected to the drone body (1); and a cleaning plate (24) is threadedly connected to the threaded rod (23).

6. A wind turbine blade damage detection device according to claim 5, characterized in that: A plurality of cleaning brushes (25) for cleaning the surface of the high-definition camera (11) are fixedly connected to the cleaning plate (24) on one side close to the high-definition camera (11); the cleaning plate (24) is slidably connected to a fixing rod (18); the upper portion of the fixing rod (18) is fixedly connected to the first connecting plate (12); and the lower portion of the fixing rod (18) is fixedly connected to the drone body (1).

7. A wind turbine blade damage detection device according to claim 1, characterized in that: The detection module comprises an ultrasonic detection unit, a visual detection unit and an infrared detection unit. The ultrasonic detection unit transmits and receives ultrasonic waves through a plurality of ultrasonic sensors (10) to detect whether there are cracks or structural defects in the wind turbine blade. The visual detection unit captures images or videos of surface damage of the wind turbine blade through a high-definition camera (11). The infrared detection unit detects the temperature distribution on the surface of the wind turbine blade through an infrared sensor (9) to perform damage detection.

8. A wind turbine blade damage detection device according to claim 7, characterized in that: The ultrasonic detection unit realizes three-dimensional positioning of damage through phased array ultrasonic detection, transmits and receives ultrasonic waves simultaneously through multiple ultrasonic probes, and processes and analyzes the information to obtain information of different depths and positions inside the blade, thereby realizing three-dimensional imaging and damage positioning.

9. A wind turbine blade damage detection device according to claim 1, characterized in that: The data fusion and analysis module uses the YOLO model to fuse the information collected by different detection units. The specific steps of the YOLO model fusing the information collected by different detection units are: Data preprocessing: Collect the raw data of wind turbine blades from different detection units, remove noise, outliers and duplicate data from the raw data, standardize the data of different detection units to the same size, and then mark them, that is, select the annotation box containing the damage and mark the damage type; YOLO model training: The preprocessed ultrasonic, visual and infrared image data are synthesized into a data set, which is then split into training set, validation set and test set in proportion. The training set is used to train the constructed YOLO model, and the validation set data is used to perform parameter optimization operations to find the optimal YOLO model. Finally, the performance of the optimal YOLO model trained is judged based on the test set data to determine the final YOLO model; Model application and deployment: The trained damage detection model is applied to the actual wind turbine blade damage detection task. By inputting new ultrasonic, visual and infrared data, the model can automatically identify the damaged area and type of the blade.

Images