Remote control type electric wire deicing unmanned aerial vehicle

Abstract

The invention discloses a remote control type electric wire deicing unmanned aerial vehicle. The unmanned aerial vehicle comprises a deicing module, a moving module, an anti-freezing module and a flying module. Wherein the deicing module is used for removing ice on the surface of a wire; the moving module is used for realizing on-line and on-line walking of the whole rack and off-line work after deicing is finished; the anti-freezing module is used for realizing protection work after deicing and spraying an anti-freezing solution to prevent the electric wire from being coated with ice again; and the flight module is used for realizing flight and landing of the whole machine. According to the unmanned aerial vehicle, the technologies of roller ice grinding, triangular anti-falling, anti-freezing spraying and wing opening and closing are creatively adopted, the design of the unmanned aerial vehicle for deicing the high-voltage line is achieved, the manpower and time cost for deicing is saved, the deicing speed and efficiency of the high-voltage line are improved, the deicing coverage is expanded, and the potential safety hazard of manual deicing is reduced.

Description

Technical Field

[0001] This invention belongs to the field of remotely controlled unmanned aerial vehicle (UAV) technology, and specifically relates to a remotely controlled wire de-icing UAV. Background Technology

[0002] Due to factors such as topography and climate, high-voltage power lines in many areas of my country, including Zhaotong in Yunnan Province and Bijie, Qiannan, and Qiandongnan in Guizhou Province, experience icing. When ice forms, wind can cause ice crystals to oscillate, leading to collisions between the ground wire and the conductor. In severe cases, this can cause short circuits and wire burnouts. As the ice thickness increases, the load on the conductors also increases, causing power poles to twist, sink, or even tilt, potentially leading to pole collapse. When the accumulated ice reaches a certain weight, the weight of the transmission lines also increases, ultimately causing flashover accidents. The natural melting of ice on high-voltage power lines is slow; therefore, manual de-icing is necessary in areas with icing. Icing areas are often located in high-altitude mountainous or hilly regions, which makes construction difficult, costly, and dangerous.

[0003] Based on this, the present invention develops a remotely controlled de-icing drone, aiming to reduce the difficulty and risk of operation and improve de-icing efficiency. Summary of the Invention

[0004] To enable on-site assessment of the overall insulation status of existing cables, this invention provides a method for evaluating the overall insulation status of existing cables based on coaxial mode velocity measurement. Without causing any damage to the transmission cable, the method measures the coaxial mode velocity of transient waves at different frequencies within the cable system and compares it with the coaxial mode velocities of brand-new cables and cables nearing retirement to determine the insulation status of the tested cable. This method not only provides a non-destructive assessment of the overall insulation status of existing cables but also significantly reduces on-site measurement time compared to the PDC measurement method. Furthermore, it greatly minimizes the impact of changes in on-site environmental conditions caused by long-duration single measurements on the measurement results.

[0005] The technical solution adopted by this invention to solve its technical problem is:

[0006] A remotely controlled wire de-icing drone, the drone comprising the following modules: a de-icing module, a mobility module, an antifreeze module, and a flight module;

[0007] The de-icing module is used to remove ice from the surface of the wires, and de-icing is achieved through mechanical polishing and gentle heating.

[0008] The mobile module is used to enable the entire machine to be mounted on the line, move along the line, and go offline after de-icing is completed;

[0009] The anti-freezing module is used for realizing the protection work after the deicing, and spraying the anti-freezing liquid to prevent the electric wire from being covered with ice again.

[0010] The flight module is used for realizing the flight and landing of the whole machine.

[0011] Preferably, the deicing module comprises a roller, and the surface of the roller is rough, can wrap the ice layer on the surface of the electric wire and apply pressure, and the ice layer is ground by rolling friction.

[0012] Preferably, the deicing module further comprises an auxiliary wheel, and the auxiliary wheel can prevent the roller from being too tight or too loose with the electric wire.

[0013] Preferably, the auxiliary wheel clamps the electric wire, and can be used for cleaning the ice residues still adhered to the electric wire after being ground.

[0014] Preferably, the deicing module further comprises a stepping motor, a screw rod, a connecting support and a connecting rod; the stepping motor drives the screw rod to rotate, so that the connecting support moves forward and backward in the axial direction along the screw rod, drives the connecting rod and the roller to make circular arc motion with the vertical shaft as the center, and realizes the opening and closing process.

[0015] Preferably, the moving module comprises a top wheel and side wheels; wherein the top wheel is a driving wheel, and the side wheels are driven wheels.

[0016] Preferably, the side wheels are arranged on both sides of the top wheel, and at least two side wheels are arranged on one side; the side wheels are in an inclined posture when the electric wire is laid, and form a triangular space with the top wheel.

[0017] Preferably, the anti-freezing module comprises an ultrasonic atomizer, which is used for preventing the electric wire from being covered with ice again after being deiced.

[0018] Preferably, the flight module comprises a wing, and the wing adopts a foldable design.

[0019] Preferably, the unmanned aerial vehicle further comprises a vision module, which is used for realizing the real-time transmission of images.

[0020] The technical advantages of the present application are as follows:

[0021] 1) The present application enables the robot to climb the high-voltage line in a remote control manner, and reduces the operation difficulty and danger.

[0022] 2) The present application solves the problem of how to prevent the deiced high-voltage line from being covered with ice again.

[0023] 3) The present application innovates a new deicing tool, which can greatly avoid damaging the high-voltage line during deicing, and even hopes to realize zero damage. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 An overall structure diagram in one embodiment of the present application;

[0025] Figure 2 A de-icing module structure diagram in one embodiment of the present application;

[0026] Figure 3 An opening state diagram of a wire clamping mechanism in the de-icing module in one embodiment of the present application;

[0027] Figure 4 A wire clamping diagram of the de-icing module in one embodiment of the present application;

[0028] Figure 5 A moving module structure diagram in one embodiment of the present application;

[0029] Figure 6 A front view of the moving module when clamping wires in one embodiment of the present application;

[0030] Figure 7 A diagram when the moving module is moving wires in one embodiment of the present application;

[0031] Figure 8 A flying module structure diagram in one embodiment of the present application;

[0032] Figure 9 A wing closed state diagram in one embodiment of the present application;

[0033] Figure 10 A wing expanded state diagram in one embodiment of the present application;

[0034] Figure 11 An anti-freezing module structure diagram in one embodiment of the present application;

[0035] Figure 12 A work flow diagram in one embodiment of the present application;

[0036] The reference signs are as follows: 1 rough surface roller, 2 first step motor, 3 screw rod, 4 first connecting support, 5 auxiliary wheel, 6 second connecting support, 7 connecting rod, 8 battery, 9 bottom plate, 10 side wheel, 11 first torsion spring, 12 second torsion spring, 13 top wheel, 14 walking motor, 15 second step motor, 16 gear, 17 rack, 18 middle rack, 19 wing arm, 20 brush, 21 ultrasonic atomizer, 22 spraying top wheel. DETAILED DESCRIPTION

[0037] The following will be described with reference to the accompanying drawings Figures 1 to 12The specific embodiments of the present application will now be described in detail with specific reference being made to the figures. While the specific embodiments of the present application are shown in the figures, it is understood that the present application can be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art.

[0038] It should be noted that certain terms have been used throughout the description and claims to refer to certain components. As one skilled in the art will appreciate, different persons can refer to the same component by different names. This description and claims should not be construed to limit the scope of the application to a specific embodiment, but rather the scope of the application should be determined by the language of the claims.

[0039] In order to facilitate the understanding of the embodiments of the present application, the following will be further explained and described with specific examples combined with the accompanying drawings, and each drawing does not constitute a limitation to the embodiments of the present application.

[0040] The present application provides a wire meticulous deicing unmanned aerial vehicle which can be remotely controlled and used for deicing. Figure 1 As shown, the whole machine includes five modules, which are deicing module, moving module, flight module, anti-freezing module and visual module.

[0041] The deicing module is used to realize the main work of the whole machine, i.e. removing the ice on the surface of the wire. Based on the ingenious combination of screw rod and truss structure and the design of polishing wheel and protection wheel, the work of polishing and deicing the ring wire and protecting the wire is realized, and the technical problems of difficult synchronization of opening and closing of the deicing wheels on both sides and easy damage of the wire during the working process of the polishing wheel are solved.

[0042] The moving module is used to realize the work of mounting the whole machine on the wire, walking on the wire and leaving the wire after deicing. Based on the characteristics of torsional spring force automatic opening and closing, the operation of mounting the wire at a long distance is realized, and the technical problem of not being able to realize remote control and automatically mounting the whole machine on the wire is solved.

[0043] The flight module is used to realize the flight and landing work of the whole machine. Based on the gear and rack transmission, the folding and unfolding of the wing are realized, and the problems of complexity and large resistance of hinge folding are solved.

[0044] The anti-freezing module is used to realize the protection work after deicing, and the anti-freezing liquid is sprayed to prevent the electric wire from being covered with ice again. The combination of the atomization of the ultrasonic atomizer and the brush realizes the function of uniformly applying the protective liquid, and solves the technical problems of waste and unevenness in the traditional application.

[0045] The visual module is used to realize the function of real-time transmission of images.

[0046] In one embodiment, as shown in Figure 2 , the deicing module is a roller grinding ice, which comprises: 1 a rough surface roller, 2 a first stepper motor, 3 a screw rod, 4 a first connecting bracket, 5 an auxiliary wheel, 6 a second connecting bracket, 7 a connecting rod, and 8 a battery. The rough surface roller 1 is connected to the second connecting bracket 6, and the auxiliary wheel 5 is connected to the first connecting bracket 4 in the same way. The second connecting bracket 6 is fixedly connected to the first connecting bracket 4 through the connecting rod 7. The screw rod 3 is connected to the first stepper motor 2 and then passes through the screw hole of the first connecting bracket 4. The battery 8 is fixedly connected to the rear side of the first connecting bracket 4. Such a connection layout can effectively realize the clamping of the deicing module to the electric wire, and the structure is relatively simple, which effectively reduces the weight. During the wiring process, the rough surface roller 1 wraps the ice layer on the surface of the electric wire and applies pressure, and the ice layer is ground by rolling friction. The opening and closing degree of the roller is controlled by the first stepper motor 2. The first stepper motor 2 drives the screw rod to rotate, so that the first connecting bracket 4 moves axially forward and backward along the screw rod 3, drives the connecting rod 7 and the roller 1 to make circular motion with the vertical shaft as the center, and realizes the opening and closing process. Specifically, referring to Figure 3 , the first stepper motor 2 drives the screw rod to rotate, the first connecting bracket 4 advances, and the two pairs of rollers open, realizing the opening of the wire clamping mechanism. Referring to Figure 2 , the first stepper motor 2 drives the screw rod to rotate, the connecting brackets 4 and 6 approach each other, the two pairs of rollers 1 and the auxiliary wheel 5 close, realizing the closing of the wire clamping mechanism. Figure 4 For the wire clamping, the auxiliary wheel 5 and the roller 1 clamp the electric wire to remove the remaining ice. The invention adopts a relatively mild deicing method of grinding wheel (rough surface roller), assisted by motor heat production. Heat production can reduce the hardness of the ice, so that the work of the grinding wheel can be more smoothly carried out. Through the combination with the control of the movement speed, the ice thickness can be reduced to the maximum extent without damaging the electric wire.

[0047] Therefore, it can not only ensure that there is enough space for the electric wire to enter and exit when the unmanned aerial vehicle is in and out of the line, but also control the roller to keep it in a tight pressing state on the electric wire. In addition, the roller is equipped with an auxiliary wheel 5 (i.e. a protective wheel) at the rear, which prevents the roller from being too tight or too loose on the electric wire; at the same time, the auxiliary wheel clamps the electric wire to clean the ice residue still adhering to the electric wire after being ground.

[0048] The design can keep close contact with the electric wire during the deicing process, and improve the deicing efficiency. Meanwhile, the auxiliary wheels can effectively avoid the situation that the roller is too tight or too loose with the electric wire, and ensure the stability and reliability of the deicing operation.

[0049] In another embodiment, the mobile module is located in the middle of the UAV, comprising: 9 bottom plate, 10 side wheels, 11 first torsion spring, 12 second torsion spring, 13 top wheels, 14 walking motor. The bottom plate 9 is connected with the shell through the second torsion spring 12, the side wheels 10 are connected with the first torsion spring 11 through bolts and then connected to the shell, the top wheels 13 are connected in series through the rod on the frame, and are matched with the walking motor 14 through the gear system. With such structure, the online and wire clamping operation of the UAV is realized more simply. Referring to Figure 5 When the force (i.e. the electric wire) contacts the horn-shaped bottom plate 9 from bottom to top during stringing, the bottom plate is opened together with the side wheels 10 under the action of the second torsion spring 12, referring to Figure 6 So that the whole machine is clamped on the electric wire from above, and then the bottom plate is reset due to the action of the second torsion spring 12, realizing wire clamping.

[0050] In another embodiment, the mobile module includes top wheels 13 and side wheels 10, two top wheels connected with 14 walking motor as driving wheels; the side wheels are driven wheels, two on each side, in a narrow-down wide-up inclined posture during wire walking, forming a triangular space with the top wheels, referring to Figure 7 So that the fuselage can keep balance on the electric wire and will not shake left and right or even hang upside down. When the whole machine flies away from the electric wire, the electric wire gives the side wheels downward force, so that the torsion spring works and the bottom plate is opened.

[0051] The mobile module is composed of one driving wheel at the top and two driven wheels on the side, which together form a triangular space to keep the UAV balanced on the electric wire. The inclined design of the driven wheels helps to stabilize the attitude of the UAV and prevent it from shaking left and right or hanging upside down. This design enables the UAV to keep balance on the electric wire, avoiding accidents caused by unstable attitude, improving operation efficiency and safety.

[0052] In another embodiment, referring to Figure 8The flight module is designed as a quadcopter, located on the upper part of the fuselage, and features a foldable design. The flight module includes: a second stepper motor (15), a gear (16), a rack (17), a central rack (18), and a wing arm (19). The second stepper motor (15) is fixed to the frame and meshes with the central rack (18) via a gear. The central rack (18) then meshes with a pair of racks (17) on either side via gear (16). The racks (17) then mesh with the wing arms (19) via gears. This design cleverly utilizes gear transmission, allowing wing opening and closing control to be achieved with a single motor. Using the second stepper motor (15) as the power source, the wing opening and closing is achieved through the transmission of gears (16), racks (17), and the central rack (18). For details, refer to... Figure 9 The second stepper motor 15 rotates counterclockwise, causing the gear 16 to rotate clockwise via the central rack 18. The gear 16 then meshes with the gear on the wing via the rack 17, causing it to rotate counterclockwise, thus closing the right wing. The left wing closes via symmetrical rack and pinion transmission. (Refer to...) Figure 10 The second stepper motor 15 rotates clockwise, causing the gear 16 to rotate counterclockwise via the central rack 18. The gear 16 then meshes with the gear on the wing via the rack 17, causing it to rotate clockwise, thus deploying the right wing. The left wing deploys via symmetrical rack and pinion transmission. The central rack 18 transmits the rotation of the vertical gear to the horizontal gears on both sides. The wing arm 19 is designed with a hollowed-out structure to achieve the wheelbase required for the UAV's payload takeoff, thus reducing weight.

[0053] The flight module features a quadcopter design. The propeller size and motor model can be determined based on the overall dimensions of the drone and lift assessment. For example, if the quadcopter drone weighs 5kg (with load), the motor model can be T-MOTOR MN4014KV400, the ESC model T-MOTOR T60A, and the propeller model T-MOTOR 16. The drone features a 5.4CF battery (LiPo 6S-22.2V-35C-8000mAh) and a 710mm wheelbase. The wings are foldable, opened and closed via a stepper motor and other power sources, while a hollow design reduces weight and improves flight efficiency. This design allows the drone to maintain balance and stability during flight maneuvers, enhancing its versatility and applicability.

[0054] In another embodiment, refer to Figure 11The anti-freezing module comprises: 20 brushes, 21 ultrasonic atomizer, 22 spraying top wheel. Among them, the front of the anti-freezing module is provided with a row of brushes 20 for cleaning the ice residues left after ice removal, the rear is provided with a spraying device of the ultrasonic atomizer 21, two rows on each side, three in each row, realizing spraying of the anti-freezing liquid on the electric wire, the spraying top wheel 22 is in contact with the electric wire, and the friction between the anti-freezing module and the electric wire is reduced by rolling, reducing the loss and effectively avoiding the reduction of the travel speed. Although the front end sanding and heating deicing can remove the ice to some extent, the treatment result is not thorough, in order to achieve the optimal deicing effect, two-sided brushes are arranged at the tail end of the application, the brushes can almost completely wrap the electric wire, and the residual ice is brushed away by means of the overall movement, so that the maximum cleaning is achieved. In addition, in order to expand the deicing result, the tail end is additionally provided with an anti-freezing liquid atomizing and spraying device, the anti-freezing liquid is sprayed on the surface of the cleaned electric wire, the risk of subsequent refreezing is reduced, and the overall benefit of deicing is increased.

[0055] The design can prolong the deicing cycle period and improve the durability of the deicing effect. The use of the anti-freezing module can effectively prevent the occurrence of re-icing after deicing, thereby reducing the number of operations, improving the operation efficiency and cost-effectiveness.

[0056] The camera is located at the bottom of the unmanned aerial vehicle. The image information collected by the camera is used as input information to control the unmanned aerial vehicle to reach the position of the electric wire by using the mature visual recognition technology, and the technology at this position can be used for reference.

[0057] In another embodiment, the visual module of the application is used to accurately identify the position of the electric wire to clamp the wire, and the front camera is used for obstacle avoidance and inspection.

[0058] Specifically, the visual identification of the up and down lines includes the following steps:

[0059] S100. A downward-looking camera is installed at the bottom of the unmanned aerial vehicle or at an inclined angle, for example, an IMU camera is selected, after real-time images are collected, the real-time images are transmitted to the main control module, the main control module converts the images into grayscale images for processing, then a Gaussian filter is used to reduce background noise and interference, and a Canny algorithm is used to extract the edge profile in the image to enhance the features of the electric wire, thereby completing the preprocessing of the real-time image.

[0060] S200. Hough transform is used to detect straight lines in the image, and all possible line segment sets are output to identify the electric wire;

[0061] S201. The Cartesian coordinates (x, y) corresponding to each edge point in the line image are converted into polar coordinate representation to calculate (p, q) values, and a peak value is found in the Hough space to determine the straight line existing in the image;

[0062] Specifically:

[0063] For each edge point (x,y), substitute it into the polar coordinate formula: ρ = x·cosθ + y·sinθ,

[0064] In order to find all possible straight lines, a series of θ values need to be sampled, for example, from 0° to 180°, with a certain step size (e.g. 1°);

[0065] For each (x,y) and each θ, calculate the corresponding ρ value;

[0066] Construct a Hough space (also known as a voting matrix), with θ as the horizontal axis and ρ as the vertical axis. Each (ρ,θ) pair corresponds to a cell. Whenever a (x,y) point falls on a straight line (ρ,θ), increase the counter (called "vote") in the corresponding cell; the more votes, the more likely the existence of this straight line.

[0067] The table is shown as follows:

[0068]

[0069] The position with a larger value in the Hough space (i.e. the (ρ,θ) combination with the most votes) represents a possible straight line in the image.

[0070] S202. Among all the detected line segments, filter out those with longer length, approximately horizontal direction (±10°) and located in the center area of the image as the wire candidates;

[0071] S203. Calculate the midpoint coordinates of multiple candidate line segments to determine the wire center position, thereby controlling the horizontal fine adjustment of the UAV;

[0072] S300. The UAV is on the line;

[0073] S301. When the offset value between the wire center position and the image center of the UAV is below a certain threshold, start the descent process until the wire is contacted;

[0074] S302. After contacting the wire, the force generated by the wire on the on-line device triggers the torsion spring to open the clamping slot, completing the on-line action;

[0075] S400. After issuing the "offline" command, the wings of the UAV are opened and lifted, and the side wheels are pushed away from the bottom plate due to the downward force of the wire, thereby detaching from the wire.

[0076] For the front camera obstacle avoidance and emergency stop, a wide-angle camera needs to be installed on the front side of the UAV to collect images in real time, zoom to the target size, improve processing speed, input the image into the target detection model, and classify and identify typical obstacles such as birds, branches, and power poles based on the YOLOv5 algorithm (YOLOv5 is an advanced algorithm for target detection based on the YOLO series framework. It uses a single-stage detection approach to quickly and accurately identify multiple targets in images or videos. The algorithm uses CSPDarknet as the backbone network to extract image features, PANet for feature fusion to combine feature information at different levels. At the same time, the anchor box mechanism is used to predict target boxes of different sizes and proportions, and an adaptive anchor box calculation method is used to adapt to different datasets) and output the obstacle class and bounding box coordinates. Then, the depth value of the target position, i.e. the actual distance to the obstacle, is obtained based on the ToF or binocular vision algorithm, the obstacle distance is compared with the set safety threshold, and when the distance is lower than the threshold, the UAV stops moving forward and sends an obstacle information. If the obstacle is cleaned in time, the UAV continues to move forward. The specific steps include:

[0077] S10. A wide-angle camera is installed on the front side of the UAV to collect images in real time, and the images are zoomed to the target size to improve processing efficiency;

[0078] S20. The zoomed image is preprocessed. First, the image size is adjusted to the required size for YOLOv5, which is commonly 640x640x3 (RGB image). Then, the pixel values are scaled to [0, 1] through normalization, i.e. img=img / 255.0. Then, the dimensions are transformed from HxWxC to CxHxW format to adapt to the PyTorch model input. Finally, the batch dimension is added to increase the batch dimension, and the final format is 1x3x640x640;

[0079] S30. The preprocessed image is input into the target detection model to detect and identify the obstacle class in front of the UAV, and output the obstacle class label and the corresponding bounding box coordinate information. The detection model mainly has three parts:

[0080] The first part is the Backbone (feature extraction) module, which uses the CSPDarknet53 backbone network to extract features of different resolutions. The input dimension is 1x3x640x640 after preprocessing, and the output feature map size (stride is 8, 16, and 32 respectively) is P3 (1x256x80x80), P4 (1x512x40x40), and P5 (1x1024x20x20);

[0081] The second part is a Neck (feature fusion) module, which adopts an improved PANet structure. The improved PANet structure up-samples high-level semantic features (P5) output by an FPN (feature pyramid network) and transversely connects and fuses the high-level semantic features with middle-level feature maps (P4). This process is repeated layer by layer upwards, and P4 is fused with P3, so as to fully integrate high-level semantic information and low-level spatial detail information, so that the model can extract more rich features when facing different size targets, which is particularly helpful for small target detection.

[0082] The third part is a Head (detection head) module. The Head module sends fused feature maps from different scales (P3, P4, P5) into a detection head for target prediction. Each scale of feature map is responsible for detecting a specific size target, and the output tensor has a dimension of 1x3x(5+C)xSxS, where 3 represents that three anchor boxes are predicted for each grid point, 5 represents that each box contains position parameters (x, y, w, h) and confidence (obj), C is the number of target categories, and SxS is the spatial size of the corresponding feature map, for example, P3 outputs 80x80, and the corresponding output dimension is 1x3x(5+80)x80x80. After decoding and screening all the detection results of different scales, the model finally outputs complete detection box results.

[0083] S40. The depth value of the obstacle is obtained by using an algorithm, and the actual distance between the obstacle and the unmanned aerial vehicle is calculated by combining the center point or the key point of the bounding box. The specific process is as follows: first, the center point coordinates (cx, cy) of each obstacle are extracted from the output of the target detection model, which will be used as the reference point for depth calculation. Here, the sensor adopts a binocular vision method, that is, the depth is calculated by using "parallax matching + triangulation". The calculation formula of the depth distance D is as follows:

[0084]

[0085] Wherein f is the focal length of the camera, B is the baseline distance (the distance between the centers of the two cameras) of the binocular camera, and d is the parallax (pixel difference) of the corresponding points in the left and right images.

[0086] In addition, in order to realize a more intelligent obstacle avoidance decision system, the obstacle is set with a category priority and a corresponding safety distance threshold. The threshold for "bird" is preliminarily set to 0.5 meters, the threshold for "branch" is preliminarily set to 1 meter, and the threshold for "electric pole" is preliminarily set to 1.5 meters. Finally, the category, center coordinates and depth distance of each obstacle are output together, which are used as the input basis for the obstacle avoidance judgment and control strategy in S50. The information is used to compare whether it is lower than the set threshold of each category, and to decide whether to trigger an emergency stop or path adjustment;

[0087] S50. On the basis of S40, if the obstacle is located on the flight path and the distance is less than the safety threshold, the obstacle avoidance response mechanism is triggered, the unmanned aerial vehicle stops advancing and sends obstacle information; if the obstacle disappears or moves out of the detection area within a set time, it continues to advance.

[0088] For power line inspection and image real-time transmission, the main purpose is to check the abnormality of the power line in the running path and return the image in real time. The host board needs to start a multi-thread mechanism to continuously collect power line pictures and cache them in an image queue. After the image is grayscale, its brightness histogram change is analyzed. Further, to reduce transmission pressure, the image is compressed in JPEG / WebP format, and then the compressed image is packaged through UDP / Socket form, with timestamp and status information. The image is uploaded to the ground station through the wireless transmission module, and the ground station uses the GUI interface to display the real-time picture.

[0089] In another embodiment,

[0090] When the present artificial remote control unmanned aerial vehicle, it can be understood that the image is transmitted to the remote controller with a display in real time, for example, the remote controller can be a tablet computer or a mobile phone directly, which directly receives the picture of the unmanned aerial vehicle camera in the existing wireless transmission protocol, and then the artificial judges whether there is an abnormal power line feature and how to continue to operate the deicing unmanned aerial vehicle through the picture.

[0091] In another embodiment,

[0092] In addition to artificial remote control of the unmanned aerial vehicle, the model of the existing YOLO algorithm (such as YOLOv5 algorithm) can be trained using the power line data set in the field to identify abnormal power line features.

[0093] For YOLO-based model detection, if abnormal power line features are detected, a warning window can be further popped up or the image can be archived, and the operator is reminded to make a manual judgment.

[0094] For example, for the model training and data set related problems of YOLOv5 algorithm, the following three aspects are explained:

[0095] Dataset construction and annotation: Due to the particularity of the application scene of the present application, it is difficult to find an existing relatively suitable dataset, so in order to target typical obstacles around the power transmission line, a special target detection dataset is constructed. First, the main adaptation data source needs to be collected by flying and shooting near the power line by a drone, including images of birds, branches, power poles and the like under different light, angle and background; second, samples matching the target categories are extracted from open datasets such as COCO and Open Images for enhanced training; then image enhancement tools can also be used to rotate, scale, blur, and simulate images, etc. Operation, improve the robustness of the model to complex scenes. Image annotation uses LabelImg tool, uses VOC or YOLO format for annotation, target categories include but are not limited to "0: bird, 1: branch, 2: pole, 3: other" and the like, each image contains an average of 1~3 targets, and the total number of samples is not less than 3000;

[0096] Model training process: YOLOv5s version is used as the basic model, PyTorch framework is used for training, the main parameters are as follows: the input size is set to 640x640, the Batch Size is set to 16, the optimizer is SGD (momentum 0.937, weight decay 0.0005), the initial learning rate is set to 0.01, the Cosine decay strategy is used, the training round is initially set to 300 rounds (adjust according to the convergence situation), the loss function includes two parts, CIoU Loss for bounding box regression, and BCE Loss for classification and confidence, and the Anchor size is regenerated by k-means clustering method to match small targets.

[0097] The training process uses 70% data for training, 20% for verification, and 10% for testing, and monitors two indicators "mAP@0.5" and "mAP@0.5:0.95", the former represents the average precision calculated when the IoU threshold is 0.5, reflecting the accuracy of the model in detecting targets under relatively loose conditions, commonly used for preliminary evaluation of model performance; represents the average precision under multiple threshold values from 0.5 to 0.95 (step size 0.05), more comprehensive measure of model detection accuracy and robustness under different strictness, is a more authoritative performance indicator.

[0098] Target optimization strategy: First, the identification weight of small targets such as "birds" and "branches" can be enhanced by weighted loss function or oversampling technique, and Focal Loss can be introduced into the classification loss to reduce the weight of simple samples such as poles, focus on difficult bird and branch targets, and alleviate the class imbalance problem. Then, the input image size is dynamically adjusted according to the distance between the unmanned aerial vehicle and the obstacle (640x640 for short distance and 832x832 for long distance), which can ensure the detection accuracy of small targets at long distance while avoiding redundant calculations at short distance.

[0099] In another embodiment, a method for remotely controlling a wire deicing unmanned aerial vehicle based on the present application, the steps include: the unmanned aerial vehicle flies to the target wire through the wings; the torsional spring of the moving module pushes the side wheels and the top wheels to clamp the wire; start deicing operation, the roller polishes the ice, the screw dynamically adjusts the pressure; the brush removes the residual ice, and the atomizer sprays the anti-freezing liquid; the walking motor drives the unmanned aerial vehicle to walk along the wire to complete continuous deicing; after deicing is completed, the moving module releases the wire and returns.

[0100] Specifically, as shown in Figure 12 , first, remotely control the unmanned aerial vehicle to take off, when the unmanned aerial vehicle is above the wire to be deiced, remotely control the unmanned aerial vehicle to descend, and perform the wire erection operation, at this time, the wire acts as an upward force, so that the bottom plate under the moving module opens inward under the action of the torsional spring, so that the wire enters the interior of the unmanned aerial vehicle, after entering, the bottom plate is reset from below under the action of the torsional spring, and the wire erection is successful; the flight module folds the wings under the gear transmission, and then the deicing module starts to work, the motor drives the screw to rotate, so that the roller covers the wire, and the deicing work is performed under the protection of the auxiliary wheel; at this time, the moving module slowly walks the line under the action of the motor; the visual module transmits the surface condition of the wire in real time during the process; after the deicing work is completed, the motor drives the screw transmission, so that the roller leaves the wire; the wings open under the action of the gear transmission; the wings work to lift the unmanned aerial vehicle, so that the wire gives the moving module side wheels and bottom plate downward force, so that it opens downward under the action of torsion, and the offline is successful; finally, remotely control it to return to the ground. The present application adopts modular design, and deicing unmanned aerial vehicle is divided into multiple relatively independent modules, which has high flexibility and adjustability.

[0101] The present application is mainly a remotely controllable high-altitude wire deicing unmanned aerial vehicle. For this research, the following solutions are proposed:

[0102] (1) For the polishing type deicing mechanism in the deicing device, different surface structures, radii, etc. can be selected according to the ice thickness of the wire and the thickness of the wire, which has good flexibility and adaptability.

[0103] The deicing mechanism in the deicing device can be appropriately increased in size according to the electric wire for the thicker electric wire to realize deicing of the thicker electric wire, and the deicing mechanism can be reduced in arc and size for the thinner electric wire to realize deicing of the thinner electric wire.

[0104] 2. The opening and closing mechanism in the deicing device, the wire clamping mechanism and the clamping mechanism in the walking mechanism can all adapt to electric wires within a certain range, and have high flexibility.

[0105] The opening and closing mechanism, the wire clamping mechanism and the clamping mechanism all have a certain range and can adapt to deicing of different electric wires, improve the flexibility of the system and reduce the development cost.

[0106] Moreover, the present scheme adopts modular design, and the deicing unmanned aerial vehicle is divided into a plurality of relatively independent modules, including a deicing module, a moving module, a flight module, a freeze-proof module and a vision module, and has high flexibility and adjustability. Each module is responsible for a part of function, and the modules interact through clear interfaces. In this way, the complexity of the system can be reduced, and the cohesion and maintainability of the system can be improved.

[0107] Compared with the conventional power line deicing tool, the present scheme has the following advantages:

[0108] First, the present scheme uses an unmanned aerial vehicle as a carrier, and combines the motion advantages of the unmanned aerial vehicle and the deicing robot.

[0109] Motion advantage one: because the power line is often located at a high altitude and connected by a power transmission tower with a height of dozens of meters, the deicing robot and the deicing trolley still need to be placed on the power line by artificial operation to perform deicing work, which obviously increases the labor cost and increases the risk of personnel work. The present scheme combines the motion advantage of the unmanned aerial vehicle, uses a flight control element and a vision module to realize remote control of the unmanned aerial vehicle, and personnel can remotely control the unmanned aerial vehicle to realize up-and-down line operation on the ground, thereby avoiding the risk of personnel climbing the tower.

[0110] Motion advantage two: because each section of power line of a certain distance needs to be connected by a tower, the conventional deicing tool is difficult to directly move to another section of power line through the power transmission tower after processing a section of power line, and still needs human assistance, which greatly reduces the deicing efficiency. The remote control and flight of the present scheme can enable the deicing unmanned aerial vehicle to fly to another section of power line at the junction of the power lines, so as to quickly continue deicing.

[0111] Third, the present application adopts the design of rotating wings to increase the stability of the deicing unmanned aerial vehicle. Because the deicing tool is directly moving on the high-voltage wire and the diameter of the power transmission line is often smaller than the deicing tool, the balance of the deicing tool in the high altitude is easily affected by the wind and shakes, which increases the risk of deicing work. In order to reduce this risk, the present application adopts a rotating wing connecting arm, which can be parallel to the space relationship with the power transmission line through rotation. On the one hand, this rotating arm design makes the weight of the whole machine concentrate near the power transmission line as much as possible, reduces the moment of the wing relative to the power transmission line, and reduces the risk of rollover. On the other hand, the wing is rotated to the two sides of the power transmission line, the wind area of the whole machine is relatively reduced, thereby the influence of the lateral force of the wind is maximally reduced, and the risk is also reduced to a certain extent. The design of the rotating arm achieved by the combination of rack and pinion can also be remotely controlled by controlling the motor to freely control the rotation of the wing arm, and finally the risk is maximally reduced.

[0112] Second, the present application adopts the deicing method of mechanical grinding wheel and heating in parallel, which is relatively mild and efficient. The traditional deicing tool generally has two ways: mechanical cutting or heating by spraying fire. Direct cutting and spraying fire can easily damage the power transmission line and thus be not worth the loss. The present application adopts the relatively mild deicing method of grinding wheel, assisted by motor heat. The heat production can reduce the hardness of the ice cover, so that the work of the grinding wheel can be more smoothly carried out, and through the combination with the movement speed control, the ice thickness can be maximally reduced without damaging the power transmission line. The combination of the two ways ensures the efficiency and safety of the deicing work, and has advantages over the traditional single deicing tool.

[0113] Third, the present application adopts the design of rotating wings to increase the stability of the deicing unmanned aerial vehicle. Because the deicing tool is directly moving on the high-voltage wire and the diameter of the power transmission line is often smaller than the deicing tool, the balance of the deicing tool in the high altitude is easily affected by the wind and shakes, which increases the risk of deicing work. In order to reduce this risk, the present application adopts a rotating wing connecting arm, which can be parallel to the space relationship with the power transmission line through rotation. On the one hand, this rotating arm design makes the weight of the whole machine concentrate near the power transmission line as much as possible, reduces the moment of the wing relative to the power transmission line, and reduces the risk of rollover. On the other hand, the wing is rotated to the two sides of the power transmission line, the wind area of the whole machine is relatively reduced, thereby the influence of the lateral force of the wind is maximally reduced, and the risk is also reduced to a certain extent. The design of the rotating arm achieved by the combination of rack and pinion can also be remotely controlled by controlling the motor to freely control the rotation of the wing arm, and finally the risk is maximally reduced.

[0114] Fourthly, the present application designs ice residue treatment device and anti-freezing coating spraying device at the tail end. Although the front end grinding and heating deicing can remove the ice to some extent, the treatment result is not complete. In order to achieve the optimal deicing effect, the tail end of the present application is provided with two-sided brushes which can almost completely wrap the power transmission line. The residual ice is removed by the movement of the whole to maximize the cleaning. In addition, in order to expand the deicing results, the tail end is additionally provided with an anti-freezing liquid atomizing spraying device to spray the anti-freezing liquid on the cleaned power transmission line surface, reduce the risk of subsequent refreezing and increase the overall benefit of deicing.

[0115] The above are the main four advantages of the present application, which are different from the conventional deicing tools.

[0116] In summary, the present application realizes the design of a high-voltage line deicing unmanned aerial vehicle by innovatively adopting the technologies of roller ice grinding, triangular anti-falling, anti-freezing spraying and wing opening and closing, saves the labor and time cost of deicing, improves the speed and efficiency of high-voltage line deicing, expands the coverage of deicing and reduces the safety hazards of manual deicing. The technologies comprehensively exhibited by the design have originality in the field of wire deicing robots.

[0117] The general description of the invention involved in the present application and the description of the specific embodiments thereof should not be understood as a limitation on the technical solutions of the invention. Based on the disclosure of the present application, the skilled in the art can add, reduce or combine the disclosed technical features in the general description or / and the specific embodiments (including examples) without violating the elements of the invention involved, to form other technical solutions within the protection scope of the present application.

Claims

1. A remote-controlled electric wire de-icing drone, characterized in that, The unmanned aerial vehicle comprises the following modules: an ice-removing module, a moving module, an anti-freezing module and a flying module. The ice-removing module is used for removing the ice on the surface of the electric wire, and the ice removal is realized by mechanical polishing and mild heating. The moving module is used for realizing the work of mounting the whole machine on the electric wire, walking on the electric wire and leaving the electric wire after the ice removal. The anti-freezing module is used for realizing the protection work after the ice removal, and spraying the anti-freezing liquid to prevent the electric wire from being covered with ice again. The flying module is used for realizing the work of flying and landing of the whole machine.

2. The drone of claim 1, wherein, Preferably, the ice-removing module comprises a roller.

3. The drone of claim 2, wherein, The surface of the roller is rough, which can wrap the ice layer on the surface of the electric wire and apply pressure, and the ice layer is ground by rolling friction.

4. The drone of claim 1, wherein, The ice-removing module comprises an auxiliary wheel.

5. The drone of claim 1, wherein, The ice-removing module comprises a connecting support and a connecting rod.

6. The drone of claim 1, wherein, The ice-removing module comprises a stepping motor.

7. The drone of claim 1, wherein, The ice-removing module comprises a screw rod.

8. The drone of claim 1, wherein, The anti-freezing module comprises an ultrasonic atomizer.

9. The drone of claim 1, wherein, The flying module comprises a wing, and the wing adopts a foldable design.

10. The drone of claim 1, wherein, The unmanned aerial vehicle further comprises a vision module, which is used for realizing the function of real-time image transmission.

Images