Electric rust removal device based on unmanned aerial vehicle and operation method

Abstract

The invention relates to an electric rust removal device based on an unmanned aerial vehicle. The electric rust removal device comprises an unmanned aerial vehicle body, a carrying platform, a steel wire brush, an unmanned aerial vehicle supporting frame and a ground control platform. A camera recognition mechanism is arranged at the top of the unmanned aerial vehicle body, a carrying platform and an unmanned aerial vehicle supporting frame are arranged at the bottom of the unmanned aerial vehicle body, the carrying platform is sleeved with the unmanned aerial vehicle supporting frame, an angle adjusting mechanism is arranged on the side face of the carrying platform, and a telescopic rod is arranged at the end of the angle adjusting mechanism. The steel wire brush is arranged at the end of the telescopic rod, the ground control platform is arranged on the ground and used for receiving information transmitted by the unmanned aerial vehicle, and an angle display module and a remote control module are arranged in the ground control platform. And a central controller is arranged on the unmanned aerial vehicle body. By optimizing the structural design and improving the rust removal scheme, the unmanned aerial vehicle can stably carry the electric steel wire brush and achieve efficient rust removal, it is ensured that the steel wire brush makes uniform contact with the target surface, and therefore the ideal rust removal effect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicles (UAVs), and more particularly to an electric rust removal device and an operation method based on UAVs. Background Art

[0002] As an important mechanical equipment, cranes play a key role in the industrial and construction fields. Their application scope covers a wide range of fields such as construction sites, ports, manufacturing, warehousing, power and energy, mining and other heavy industries. With the large-scale use of cranes, their safety and performance maintenance have become crucial issues. In this context, the formation and spread of rust has become a major hidden danger affecting the operation of cranes. Rust is not only an aesthetic problem on the surface of the crane, but more importantly, it may cause damage to the mechanical structure, affecting the stability and working performance of the crane.

[0003] As an important grinding tool, wire brushes are often used for workpiece surface treatment, such as paint removal, rust removal, and burr removal. The application scope is basically manual hand-held angle grinders equipped with wire brushes for grinding. Manual grinding is inefficient, the quality is uneven, and there are certain risks. At present, the rust removal of steel structures can be completed by drones, which use drones as carriers of electric wire brushes and use the movement and hovering capabilities of drones to complete the metal rust removal of different structures. In the prior art, rust removal work in complex high-altitude environments has the problems of high difficulty in operation, high safety risks, high investment in manpower and material resources, and low efficiency. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide an electric rust removal device and operation method based on a drone, which greatly improves the efficiency and effect of rust removal while reducing the difficulty of operation and safety risks, and provides a reliable solution for rust removal operations in complex high-altitude environments.

[0005] The technical solution adopted by the present invention to solve the technical problem is: construct an electric rust removal device based on a drone, including a drone body, a carrying platform, a wire brush, a drone support frame and a ground control platform; a camera recognition mechanism is arranged on the top of the drone body, a carrying platform and a drone support frame are arranged on the bottom of the drone body, the drone support frame is sleeved on the outside of the carrying platform, an angle adjustment mechanism is arranged on the side of the carrying platform, a telescopic rod is arranged at the end of the angle adjustment device, the wire brush is arranged at the end of the telescopic rod, and the ground control platform is arranged on the ground and is used to receive information transmitted by the drone;

[0006] A placement cavity is arranged inside the carrying platform, a telescopic rod controller and a power supply are arranged inside the placement cavity, the telescopic rod controller and the power supply are electrically connected, a force feedback mechanism is arranged on the surface of the wire brush, a force sensor is arranged on the surface of the force feedback mechanism, an electric rotating shaft is arranged on the carrying platform, and the electric rotating shaft is connected to the bottom of the telescopic rod;

[0007] An angle display module and a remote control module are arranged in the ground control platform; and a central controller is arranged on the drone body.

[0008] According to the above scheme, the ground control platform is used to receive information transmitted by the UAV.

[0009] According to the above scheme, the angle display module displays the current angle of the telescopic rod in real time, helping the operator to make precise adjustments through the remote control; the remote control module receives the operator's angle adjustment instructions through wireless communication, controls the movement of the electric rotating shaft, and realizes manual adjustment of the telescopic rod angle.

[0010] According to the above scheme, the central controller is used to process the data signals of the camera recognition mechanism, the force feedback mechanism and the force sensor, and transmit the analysis result instructions to the telescopic rod controller to achieve control and adjustment.

[0011] According to the above scheme, the force sensor is used to detect the force value when the wire brush contacts the target surface, and transmit the force value data to the wire brush control module;

[0012] The wire brush control module analyzes the current detection force value according to the set optimal rust removal force threshold:

[0013] If the force is too large: control the extension of the telescopic mechanism to reduce the contact pressure between the brush head and the surface to avoid damaging the target surface;

[0014] If the force value is too small: control the telescopic mechanism to shorten, increase the pressure of the brush head, and ensure that the rust spots are thoroughly cleaned;

[0015] After the extension and retraction adjustment is completed, the force sensor monitors the new force value in real time and provides secondary feedback until the optimal pressure range is reached.

[0016] According to the above solution, the electric rotating shaft is used to drive the telescopic rod to rotate in the horizontal and vertical directions to adjust the angle of the telescopic rod.

[0017] The present invention also provides an operation method based on the UAV electric rust removal device, comprising the following steps:

[0018] S1. Fix the carrying platform on the drone support frame, and then firmly install both sides of the drone support frame to the drone bracket. Control the drone to take off through remote control commands. The drone carries a wire brush and flies to the target crane. With the help of the camera recognition module, the drone collects image information on the surface of the crane and transmits the collected image to the ground control platform in real time. After the mission is completed, the drone returns to a safe area for landing;

[0019] S2. The ground control platform receives and processes the image data, uses the YOLOV8 algorithm based on deep learning target detection technology to analyze the image, and accurately identifies and quantifies the scope and degree of corrosion on the crane surface;

[0020] S3. According to the rusted area identified in step S2, the drone is controlled to fly to the target position; after the drone arrives at the operating area, the telescopic rod is adjusted by remote control to gradually extend the telescopic rod until the wire brush contacts the crane surface and stops; the angle of the telescopic rod is adjusted using the remote control, and the telescopic rod is adjusted to the optimal contact angle of the target surface in combination with the angle information displayed in real time by the angle display device. At this time, the contact pressure between the wire brush and the surface is monitored in real time through the built-in force sensor, and the currently detected force value is analyzed and adjusted: if the force value is too large, the telescopic mechanism is controlled to be further extended to reduce the contact pressure between the brush head and the target surface to avoid damage to the target surface; if the force value is too small, the telescopic mechanism is controlled to be appropriately shortened to increase the brush head pressure to ensure that the rusted area can be thoroughly cleaned. After the initial adjustment of the telescopic rod pressure is completed, the force sensor continuously monitors and feeds back the new pressure value in real time until the contact pressure is in the optimal range. After the adjustment is completed, the electric wire brush is started to start the rust removal operation.

[0021] S4, repeat the pressure adjustment and rust removal process of step S3, and repeat the above operation three to four times until the rust removal task of the target area is completed.

[0022] 8. The operation method based on the UAV electric rust removal device according to claim 8 is characterized in that, in the step S2, the image is analyzed using the YOLO V8 algorithm, which specifically includes the following steps: the YOLOv8 algorithm performs target detection through a network architecture, and the architecture includes three parts: Backbone, Neck and Head;

[0023] Backbone is used to extract the features of the image, and its calculation method is as follows:

[0024] F i+1 =ReLU(W i *F i +b i )

[0025] Among them, F i is the input feature, Wi and b i are weight and bias parameters respectively. The Neck part fuses multi-scale features through the feature pyramid network (FPN), and its formula is as follows:

[0026] P l =UpSample(C l+1 )+C l

[0027] Among them, C l and C l+1 are the lth and l+1th layer features respectively, P l is the fused feature map.

[0028] In the detection stage, YOLOv8 represents the detection box by predicting the center point coordinates (x, y)(x, y), width w and height h of the detection box:

[0029] B=(x,y,w,h)

[0030] And calculate the confidence S of the detection result, the formula is:

[0031] S=C*P(c|B)

[0032] Among them, C is the target confidence, P(c|B) is the category probability;

[0033] To optimize detection performance, YOLOv8 defines a loss function for comprehensive classification, positioning, and confidence, which is expressed as follows:

[0034] L=λ cls ·L cls +λ box ·L box +λ obj ·L obj

[0035] Among them, L cls represents the classification loss, L box Represents the positioning loss (calculated using CIOU loss):

[0036] L box =1-CIOU(B,B gt )

[0037] L obj represents the confidence loss; λ cls , box , obj is the loss weight parameter.

[0038] In the inference stage, YOLOv8 uses non-maximum suppression (NMS) to remove redundant detection boxes. The specific calculation method is:

[0039]

[0040] Among them, IoU is the intersection over union ratio of the detection box and the real box, which is used to measure the degree of overlap between the two.

[0041] The implementation of the electric rust removal device and operation method based on drone of the present invention has the following beneficial effects:

[0042] 1. The present invention uses the wire brush carried by the drone and its telescopic mechanism, combined with force feedback control technology, to accurately adjust the contact force between the wire brush and the target surface, ensuring that the rust spots are completely removed without damaging the surface. This adaptive adjustment mechanism avoids errors in manual operation and greatly improves the rust removal efficiency and accuracy. It is particularly suitable for rust removal work on large equipment such as cranes and steel structures;

[0043] 2. The present invention monitors the contact force between the wire brush and the target surface in real time through a force sensor and transmits data to the control system. The control system automatically adjusts the length of the telescopic rod and the pressure of the wire brush according to the set optimal rust removal force threshold (such as 10N to 15N) to ensure that the rust removal force can effectively remove rust spots without causing damage to the target surface. This dynamic feedback adjustment mechanism greatly improves work safety and surface protection effects, and avoids scratches or wear that may be caused by traditional methods;

[0044] 3. The camera recognition system of the present invention is combined with deep learning target detection technology to collect high-definition images and automatically analyze the target surface, accurately identify and quantify the scope and severity of the rusted area. The application of this technology makes rust removal more accurate, and maintenance personnel can obtain rust reports in real time, effectively guiding subsequent maintenance and protection work;

[0045] 4. The present invention enhances the stability of the UAV by installing a support frame, ensuring that the rust removal device can maintain a stable posture and work efficiently during flight and operation. In addition, the backup battery and quick-change module can monitor the power status in real time and issue a low-power alarm in time to ensure that the equipment is always in a state of sufficient power during long-term operation, avoiding interruption of operation due to insufficient power, and significantly improving the continuity and efficiency of the operation;

[0046] 5. The ground control platform of the present invention can receive various sensor data, image information and operation feedback from the UAV, realize remote control and monitoring of the UAV, and adjust the operation strategy in real time through data processing and analysis of the ground platform, ensure accurate control and data recording during the rust removal process, help operators to obtain maintenance progress and status in time, and improve overall operation efficiency and operation quality;

[0047] 6. The system of the present invention adopts a modular design. The wire brush module can be replaced according to different operating requirements to adapt to the cleaning requirements of different surfaces, shapes and degrees of rust. This design not only enhances the adaptability of the equipment, but also facilitates the maintenance and upgrading of the equipment, and can meet the rust removal needs under different environments and working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0049] Figure 1 It is a structural schematic diagram of an electric rust removal device based on a drone of the present invention;

[0050] Figure 2 It is a schematic flow chart of the operation method of the UAV electric rust removal device of the present invention;

[0051] In the figure: 1. UAV body, 2. carrying platform, 3. UAV support frame, 4. camera recognition mechanism, 5. telescopic rod, 6. wire brush, 7. angle adjustment mechanism, 8. ground control platform, 9. placement cavity, 10. telescopic controller, 11. power supply, 12. force feedback mechanism, 13. force sensor, 14. electric rotating axis, 15. angle display module, 16. remote control module, 17. central controller. DETAILED DESCRIPTION

[0052] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described in detail with reference to the accompanying drawings.

[0053] like Figure 1 As shown, the electric rust removal device based on a drone of the present invention includes a drone body 1, a carrying platform 2, a wire brush 6, a drone support frame 3 and a ground control platform 8. The drone body 1 is an energized flying structure, which provides flight power and plays the role of carrying the carrying platform 2 and the wire brush 6 structure to fly to the operation point. A camera recognition mechanism 4 is arranged on the top of the drone body 1 to perform camera recognition on the rusted parts of the crane. The carrying platform 2 and the drone support frame 3 are arranged at the bottom of the drone body 1. The drone support frame 3 is located on both sides of the carrying platform 2 and is connected to the drone body 1, playing a fixed support role, and at the same time providing additional stability for the drone during operation.

[0054] In a preferred embodiment of the present invention, the UAV support frame 3 is sleeved on the outside of the carrying platform 2, a placement cavity 9 is arranged inside the carrying platform 2, a telescopic rod controller 10 and a power supply 11 are arranged inside the placement cavity 9, the telescopic rod controller 10 and the power supply 11 are electrically connected, a force feedback mechanism 12 is arranged on the surface of the wire brush 6, a force sensor 13 is arranged on the surface of the force feedback mechanism 12, an electric rotating shaft 14 is arranged on the carrying platform 2, and the electric rotating shaft 14 is connected to the bottom of the telescopic rod 5. The wire brush 6 is connected to the telescopic rod 5, and the wire brush 6 is installed through the telescopic mechanism to perform rust removal tasks. The wire brush 6 is made of high-strength wear-resistant material and can cope with different degrees of rust cleaning needs. The wire brush 6 is replaceable to meet different types of operation requirements. The force feedback mechanism 12 arranged on the surface of the wire brush 6 enables the telescopic rod controller 10 to control the telescopic rod 5 to extend and retract according to the magnitude of the force.

[0055] In a preferred embodiment of the present invention, an angle adjustment mechanism 7 is provided on the side of the carrying platform 2, a telescopic rod 5 is provided at the end of the angle adjustment mechanism 7, a wire brush 6 is provided at the end of the telescopic rod 5, and a ground control platform 8 is provided on the ground and is used to receive information transmitted by the drone. The telescopic rod 5 is connected to the carrying platform 2 and is extended and retracted under the control of the telescopic rod controller 10. The angle adjustment mechanism 7 provided at the connection between the telescopic rod 5 and the carrying platform 2 is used to realize the flexible adjustment of the telescopic rod 5 within a multi-angle range. The angle adjustment mechanism 7 includes an electric rotating shaft 14, an angle adjustment mechanism 7 and a remote control module 16, and the angle display module 15 and the remote control module 16 are provided in the ground control platform 8. The electric rotating shaft 14 is installed on the carrying platform 2 and connected to the bottom of the telescopic rod 5. The electric rotating shaft 14 is used to drive the telescopic rod 5 to rotate in the horizontal and vertical directions to adjust the angle of the telescopic rod 5. The angle adjustment mechanism 7 is provided on the ground control platform 8, and the current angle of the telescopic rod 5 is displayed in real time, helping the operator to make precise adjustments through the remote control. The remote control module 16 receives the angle adjustment instruction of the operator through wireless communication, controls the action of the electric rotating shaft 14, and realizes the manual adjustment of the angle of the telescopic rod 5. The electric rotating shaft 14 is used to drive the telescopic rod 5 to rotate in the horizontal and vertical directions to adjust the angle of the telescopic rod 5 .

[0056] In a preferred embodiment of the present invention, a central controller 17 is provided on the drone body 1. The central controller 17 is used to process the data signals of the camera recognition mechanism 4, the force feedback mechanism 12 and the force sensor 13, and transmit the analysis result command to the telescopic rod controller 10 to achieve precise control and real-time adjustment. The drone body 1 is equipped with a backup battery and a quick replacement module, which monitors the power in real time and issues an alarm when the power is lower than a preset threshold, prompting ground personnel to replace it, ensuring that the drone can operate continuously for a long time and improving the operating efficiency. The ground control platform 8 is used to receive various information uploaded from the drone.

[0057] In a preferred embodiment of the present invention, the force sensor 13 is used to detect the force value when the wire brush 6 contacts the target surface, and transmits the force value data to the wire brush 6 control module; the wire brush 6 control module analyzes the current detected force value according to the set optimal rust removal force threshold: if the force value is too large: control the extension of the telescopic mechanism to reduce the contact pressure between the brush head and the surface to avoid damaging the target surface; if the force value is too small: control the shortening of the telescopic mechanism to increase the pressure of the brush head to ensure that the rust spots are thoroughly cleaned; after the telescopic adjustment is completed, the force sensor 13 monitors the new force value in real time and performs secondary feedback until the optimal pressure range is reached.

[0058] The present invention also provides an operation method based on the UAV electric rust removal device, comprising the following steps:

[0059] S1. Fix the carrying platform 2 on the drone support frame 3, and then firmly install the two sides of the drone support frame 3 to the drone bracket. Control the drone to take off through remote control commands. The drone carries a wire brush 6 and flies to the target crane. With the help of the camera recognition module, the drone collects image information on the surface of the crane and transmits the collected image to the ground control platform 8 in real time. After the mission is completed, the drone returns to a safe area for landing;

[0060] S2, the ground control platform 8 receives and processes the image data, analyzes the image using the YOLO V8 algorithm based on deep learning target detection technology, and accurately identifies and quantifies the scope and degree of corrosion of the crane surface;

[0061] The image is analyzed using the YOLO V8 algorithm, specifically including the following steps: the YOLOv8 algorithm performs target detection through a network architecture, and the architecture includes three parts: Backbone, Neck, and Head;

[0062] Backbone is used to extract the features of the image, and its calculation method is as follows:

[0063] F i+1 =ReLU(W i *F i +b i )

[0064] Among them, F i is the input feature, W i and b i are weight and bias parameters respectively. The Neck part fuses multi-scale features through the feature pyramid network (FPN), and its formula is as follows:

[0065] P l =UpSample(C l+1 )+C l

[0066] Among them, C l and C l+1 are the lth and l+1th layer features respectively, P l is the fused feature map.

[0067] In the detection stage, YOLOv8 represents the detection box by predicting the center point coordinates (x, y)(x, y), width w and height h of the detection box:

[0068] B=(x,y,w,h)

[0069] And calculate the confidence S of the detection result, the formula is:

[0070] S=C*P(c|B)

[0071] Among them, C is the target confidence, P(c|B) is the category probability;

[0072] To optimize detection performance, YOLOv8 defines a loss function for comprehensive classification, positioning, and confidence, which is expressed as follows:

[0073] L=λ cls ·L cls +λ box ·L box +λ obj ·L obj

[0074] Among them, L cls represents the classification loss, L box Represents the positioning loss (calculated using CIOU loss):

[0075] L box =1-CIOU(B,B gt )

[0076] L obj represents the confidence loss; λ cls , box , obj is the loss weight parameter.

[0077] In the inference stage, YOLOv8 uses non-maximum suppression (NMS) to remove redundant detection boxes. The specific calculation method is:

[0078]

[0079] Among them, IoU is the intersection over union ratio of the detection box and the real box, which is used to measure the degree of overlap between the two.

[0080] Based on the above algorithm, the high-definition images taken by the drone can be automatically analyzed by the ground control platform 8, so as to accurately identify the rusted area and generate an analysis report including the scope and severity of the rust, providing a reliable basis for subsequent maintenance and anti-rust treatment.

[0081] S3. According to the rusted area identified in step S2, the drone is controlled to fly to the target position; after the drone arrives at the operating area, the telescopic rod 5 is remotely adjusted to gradually extend the telescopic rod 5 until the wire brush 6 stops contacting the crane surface, and the angle of the telescopic rod 5 is adjusted using the remote control. Combined with the angle information displayed in real time by the angle display device, the optimal contact angle of the telescopic rod 5 to the target surface is adjusted. At this time, the contact pressure between the wire brush 6 and the surface is monitored in real time through the built-in force sensor 13, and the currently detected force value is analyzed and adjusted: if the force value is too large, the telescopic mechanism is controlled to be further extended to reduce the contact pressure between the brush head and the target surface to avoid damage to the target surface; if the force value is too small, the telescopic mechanism is controlled to be appropriately shortened to increase the brush head pressure to ensure that the rusted area can be thoroughly cleaned. After completing the preliminary adjustment of the pressure of the telescopic rod 5, the force sensor 13 continuously monitors and feeds back the new pressure value in real time until the contact pressure is in the optimal range. After the adjustment is completed, the electric wire brush 6 is started to start the rust removal operation.

[0082] S4: Repeat the pressure adjustment and rust removal process of step S3. The above operation is repeated three to four times until the rust removal task of the target area is completed.

[0083] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. An electric rust removal device based on a drone, characterized in that: It comprises a drone body, a carrying platform, a wire brush, a drone support frame and a ground control platform; a camera recognition mechanism is arranged on the top of the drone body, a carrying platform and a drone support frame are arranged on the bottom of the drone body, the drone support frame is sleeved on the outside of the carrying platform, an angle adjustment mechanism is arranged on the side of the carrying platform, a telescopic rod is arranged at the end of the angle adjustment device, the wire brush is arranged at the end of the telescopic rod, and the ground control platform is arranged on the ground and is used to receive information transmitted by the drone; A placement cavity is arranged inside the carrying platform, a telescopic rod controller and a power supply are arranged inside the placement cavity, the telescopic rod controller and the power supply are electrically connected, a force feedback mechanism is arranged on the surface of the wire brush, a force sensor is arranged on the surface of the force feedback mechanism, an electric rotating shaft is arranged on the carrying platform, and the electric rotating shaft is connected to the bottom of the telescopic rod; An angle display module and a remote control module are arranged in the ground control platform; and a central controller is arranged on the drone body.

2. The electric rust removal device based on drone according to claim 1 is characterized in that: The ground control platform is used to receive information transmitted by the UAV.

3. The electric rust removal device based on drone according to claim 1 is characterized in that: The angle display module displays the current angle of the telescopic rod in real time, helping the operator to make precise adjustments through the remote control; the remote control module receives the operator's angle adjustment instructions through wireless communication, controls the movement of the electric rotating shaft, and realizes manual adjustment of the telescopic rod angle.

4. The electric rust removal device based on drone according to claim 1 is characterized in that: The central controller is used to process data signals of the camera recognition mechanism, the force feedback mechanism and the force sensor, and transmit the analysis result instructions to the telescopic rod controller to achieve control and adjustment.

5. The electric rust removal device based on drone according to claim 4 is characterized in that: The force sensor is used to detect the force value when the wire brush contacts the target surface, and transmit the force value data to the wire brush control module; The wire brush control module analyzes the current detection force value according to the set optimal rust removal force threshold: If the force is too large: control the extension of the telescopic mechanism to reduce the contact pressure between the brush head and the surface to avoid damaging the target surface; If the force value is too small: control the telescopic mechanism to shorten, increase the pressure of the brush head, and ensure that the rust spots are thoroughly cleaned; After the extension and retraction adjustment is completed, the force sensor monitors the new force value in real time and provides secondary feedback until the optimal pressure range is reached.

6. The electric rust removal device and operation method based on drone according to claim 1 is characterized in that: The electric rotating shaft is used to drive the telescopic rod to rotate in the horizontal and vertical directions to adjust the angle of the telescopic rod.

7. An operation method based on an electric rust removal device of a drone, characterized in that: The following steps are involved: S1: Fix the carrying platform on the drone support frame, and then firmly install both sides of the drone support frame to the drone bracket. Control the drone to take off through remote control commands. The drone carries a wire brush and flies to the target crane. With the help of the camera recognition module, the drone collects image information on the surface of the crane and transmits the collected images to the ground control platform in real time. After the mission is completed, the drone returns to a safe area for landing; S2: The ground control platform receives and processes the image data, uses the YOLO V8 algorithm based on deep learning target detection technology to analyze the image, and accurately identifies and quantifies the scope and degree of corrosion on the crane surface; S3: According to the rusted area identified in step S2, the drone is controlled to fly to the target position; after the drone arrives at the working area, the telescopic rod is adjusted by remote control to gradually extend the telescopic rod until the wire brush contacts the surface of the crane and stops, and the angle of the telescopic rod is adjusted by the remote control, and the telescopic rod is adjusted to the optimal contact angle of the target surface in combination with the angle information displayed in real time by the angle display device. At this time, the contact pressure between the wire brush and the surface is monitored in real time through the built-in force sensor, and the currently detected force value is analyzed and adjusted: if the force value is too large, the telescopic mechanism is controlled to be further extended to reduce the contact pressure between the brush head and the target surface to avoid damage to the target surface; if the force value is too small, the telescopic mechanism is controlled to be appropriately shortened to increase the brush head pressure to ensure that the rusted area can be thoroughly cleaned. After the initial adjustment of the telescopic rod pressure is completed, the force sensor continuously monitors and feeds back the new pressure value in real time until the contact pressure is in the optimal range. After the adjustment is completed, the electric wire brush is started to start the rust removal operation. S4: Repeat the pressure adjustment and rust removal process of step S3. The above operation is repeated three to four times until the rust removal task of the target area is completed.

8. The operation method based on the UAV electric rust removal device according to claim 8 is characterized in that: In the step S2, the image is analyzed using the YOLO V8 algorithm, which specifically includes the following steps: the YOLOv8 algorithm performs target detection through a network architecture, and the architecture includes three parts: Backbone, Neck and Head; Backbone is used to extract the features of the image, and its calculation method is as follows: F i+1 =ReLU(W i *F i +b i ) Among them, F i is the input feature, W i and b i are weight and bias parameters respectively. The Neck part fuses multi-scale features through the feature pyramid network (FPN), and its formula is as follows: P l =UpSample(C l+1 )+C l Among them, C l and C l+1 are the lth and l+1th layer features respectively, P l is the fused feature map. In the detection stage, YOLOv8 represents the detection box by predicting the center point coordinates (x, y)(x, y), width w and height h of the detection box: B=(x,y,w,h) And calculate the confidence S of the detection result, the formula is: S=C*P(c|B) Among them, C is the target confidence, P(c|B) is the category probability; To optimize detection performance, YOLOv8 defines a loss function for comprehensive classification, positioning, and confidence, which is expressed as follows: L=λ cls ·L cls +λ box ·L box +λ obj ·L obj Among them, L cls represents the classification loss, L box Represents the positioning loss (calculated using CIOU loss): L box =1-CIOU(B,B gt ) L obj represents the confidence loss; λ cls , box , obj is the loss weight parameter. In the inference stage, YOLOv8 uses non-maximum suppression (NMS) to remove redundant detection boxes. The specific calculation method is: Among them, IoU is the intersection over union ratio of the detection box and the real box, which is used to measure the degree of overlap between the two.

Images