A wind power blade detection method, device and equipment based on unmanned aerial vehicle tracking and medium

Abstract

The present application relates to a kind of wind power blade detection method, device, equipment and medium based on unmanned aerial vehicle tracking, belong to unmanned aerial vehicle detection field.Therein, the method includes preset wind power blade detection initial information;Preset unmanned aerial vehicle working point, unmanned aerial vehicle moves to the unmanned aerial vehicle working point based on the wind power blade detection initial information and obtains unmanned aerial vehicle working distance;Unmanned aerial vehicle carries out wind power blade target identification to wind power blade by wind power blade detection device and obtains target wind power blade image, and the wind power blade target identification is realized by YOLOv8 target detection algorithm;Based on the target wind power blade image, generate unmanned aerial vehicle tracking path by improving path generation method;Unmanned aerial vehicle carries out wind power blade defect detection according to the unmanned aerial vehicle tracking path based on PPO algorithm, realizes not to stop machine comprehensive real-time detection wind power blade defect, reduces manpower cost, improves detection efficiency, and reduces the risk of misjudgment.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of unmanned aerial vehicle detection, and particularly relates to a wind power blade detection method, device, equipment and medium based on unmanned aerial vehicle tracking. BACKGROUND

[0002] With increasing concerns about climate change and environmental pollution, wind energy is favored as a clean and environmentally friendly energy form. Converting wind energy into electricity through wind turbines is one of the common energy conversion methods. Wind energy technology has made great progress in the past few decades. The continuous innovation and development of wind energy technology have made wind power generation more competitive compared to traditional energy sources. The cost reduction is mainly due to the technical improvement of wind turbine, the improvement of manufacturing efficiency and the advantage of large-scale production. In order to make full use of wind power resources, more and more wind turbines have been built.

[0003] For wind power generation, wind power blades are the core components of wind power generation systems, directly affecting the efficiency and reliability of wind power generation. Therefore, it is extremely important to ensure the normal operation of wind power blades during wind power generation. Wind power blades are exposed to the outside for a long time due to their special nature of work, and are easily corroded and damaged. If damaged blades are not repaired in time, it is difficult to ensure the safe operation of wind power generation systems. Therefore, wind power blade inspection is an important part of wind power generation. The current wind power blade inspection method is mainly manual, which requires the wind turbine to be shut down first and then manually inspected. This method undoubtedly reduces the efficiency of wind power generation and cannot ensure comprehensive detection of wind power blades. In addition, this method cannot detect the state of wind power blades in real time, and frequent start-stop operations also affect the service life of wind turbines. SUMMARY

[0004] To solve the above problems in the prior art, the application provides a wind power blade detection method, device, equipment and medium based on unmanned aerial vehicle tracking.

[0005] The object of the application can be achieved by the following technical solutions:

[0006] A wind power blade detection method based on unmanned aerial vehicle tracking, the implementation of the wind power blade detection method includes the following steps:

[0007] S1: presetting wind power blade detection initial information, the wind power blade detection initial information including wind power blade initial information and unmanned aerial vehicle initial information;

[0008] S2: presetting an unmanned aerial vehicle working point, the unmanned aerial vehicle moving to the unmanned aerial vehicle working point based on the wind power blade detection initial information and obtaining an unmanned aerial vehicle working distance;

[0009] S3: The unmanned aerial vehicle detects the wind turbine blade through the wind turbine blade detection device to identify the target wind turbine blade and obtain a target wind turbine blade image, and the target wind turbine blade identification is realized through a YOLOv8 target detection algorithm;

[0010] S4: An improved path generation method is used to generate an unmanned aerial vehicle tracking path based on the target wind turbine blade image;

[0011] S5: The unmanned aerial vehicle detects defects of the wind turbine blade based on a PPO algorithm according to the unmanned aerial vehicle tracking path.

[0012] Preferably, the initial information of the wind turbine blade specifically includes a wind turbine blade rotating speed and a wind turbine blade rotating center position, and the initial information of the unmanned aerial vehicle includes an initial position of the unmanned aerial vehicle and an initial detection distance, the initial position of the unmanned aerial vehicle is directly in front of the wind turbine blade rotating center position, and the initial detection distance is the distance between the wind turbine blade rotating center position and the initial position of the unmanned aerial vehicle.

[0013] Preferably, the working distance of the unmanned aerial vehicle is the distance between the working point of the unmanned aerial vehicle and the wind turbine blade rotating center position, the unmanned aerial vehicle is equipped with the wind turbine blade detection device, the wind turbine blade detection device includes a laser radar and an industrial camera, and the laser radar is used to obtain the working distance of the unmanned aerial vehicle.

[0014] Preferably, the step S4 specifically includes:

[0015] S401: The Canny edge detection algorithm is used to detect the wind turbine blade edge of the target wind turbine blade image and obtain a target wind turbine blade boundary image.

[0016] S402: The unmanned aerial vehicle tracking path is generated based on the target wind turbine blade boundary image.

[0017] Preferably, the step S402 specifically includes:

[0018] S402-1: Special boundary points of the target wind turbine blade boundary image are extracted, the special boundary points are single wind turbine blade top end points, and the special boundary points include a first special boundary point, a second special boundary point and a third special boundary point.

[0019] S402-2: The special boundary points are connected to obtain a special boundary point polygon, two edges of the special boundary point polygon are randomly selected to obtain a polygon center line, and a target wind turbine blade boundary image barycenter is obtained through the intersection of the polygon center line.

[0020] S402-3: Obtain a wind turbine blade detection path set, the wind turbine blade detection path set comprising a first wind turbine blade detection path, a second wind turbine blade detection path, a third wind turbine blade detection path, and a fourth wind turbine blade detection path, the first wind turbine blade detection path being a band direction path connected from the target wind turbine blade boundary graph barycenter to the first special boundary point, the second wind turbine blade detection path being a band direction path connected from the first special boundary point to the second special boundary point, the third wind turbine blade detection path being a band direction path connected from the second special boundary point to the target wind turbine blade boundary graph barycenter, and the fourth wind turbine blade detection path being a band direction path connected from the target wind turbine blade boundary graph barycenter to the third special boundary point, and the UAV tracking path being obtained through the wind turbine blade detection path set.

[0021] Preferably, the step S5 specifically comprises:

[0022] S501: Discretize the UAV tracking path to obtain N UAV tracking path discrete points with equal distances;

[0023] S502: The UAV takes the target wind turbine blade boundary graph barycenter as a detection starting point and traverses the UAV tracking path discrete points according to the PPO algorithm;

[0024] S503: The UAV performs wind turbine blade defect detection based on the Simple Net defect detection algorithm.

[0025] Preferably, the step S502 specifically comprises:

[0026] S502-1: Pre-set a shooting period, and obtain a wind turbine blade image set through the industrial camera according to the shooting period, the wind turbine blade image set being a set of wind turbine blade images;

[0027] S502-2: Obtain a wind turbine blade image center point of the wind turbine blade image, pre-set a discrete point distance threshold and obtain a target distance, the target distance being the distance between the wind turbine blade image center point and the UAV tracking path discrete point, and when the target distance is less than or equal to the discrete point distance threshold, set the UAV tracking path discrete point as a next target point, the next target point being the UAV tracking path discrete point to be reached by the UAV in the next step;

[0028] S502-3: traversing the UAV tracking path discrete points based on the PPO algorithm to obtain a minimum discrete distance, the minimum discrete distance being a distance between the UAV and the nearest UAV tracking path discrete point, and obtaining an algorithm input state, the algorithm input state being an input state of the PPO algorithm at time t, the algorithm input state including a working distance difference value, the target distance, a discrete point position, a discrete point distance, and the wind turbine blade speed, the working distance difference value being a difference between the UAV working distance and the ideal UAV working distance, the discrete point position being a position of the UAV tracking path discrete point relative to a wind turbine blade image center point of the wind turbine blade image, the discrete point distance being a distance between the UAV tracking path discrete point and the target wind turbine blade boundary graph barycenter, and an expression of the algorithm input state being:

[0029] s t =(Dis t ,dis t ,θ t ,Δ t ,ω)

[0030] Dis t =min(Radar t )-ΔD

[0031]

[0032]

[0033] where s t is the algorithm input state, Dis t is the working distance difference value, dis t is the target distance, θ t is the discrete point position, Δ t is the discrete point distance, ω is the wind turbine blade speed, min(Radar t ) is the minimum discrete distance, ΔD is the UAV working distance, x j is a special boundary point horizontal coordinate, y j is a special boundary point vertical coordinate, x i is a UAV tracking path discrete point horizontal coordinate, y i is a UAV tracking path discrete point vertical coordinate, x0 is a target wind turbine blade boundary graph barycenter horizontal coordinate, and y0 is a target wind turbine blade boundary graph barycenter vertical coordinate.

[0034] S502-4: obtaining a reward value of the PPO algorithm and outputting UAV tracking path adjustment information based on the reward value, the UAV tracking path adjustment information including UAV tracking path non-adjustment information, UAV tracking path abnormal information, and UAV tracking path adjustment information, and an expression of the reward value being:

[0035]

[0036] wherein, rt is a reward value, Dis t is a working distance difference value, ε is a discrete point distance threshold value, Tanh is a hyperbolic tangent, dis t is a target distance, θ t is a discrete point azimuth, min(Radar t ) is a minimum discrete distance, ΔD is a working distance of the unmanned aerial vehicle;

[0037] judging the reward value, when the reward value is 100, feeding back information that the unmanned aerial vehicle tracking path does not need to be adjusted;

[0038] when the reward value is in the open interval (-100, 100), feeding back abnormal information of the unmanned aerial vehicle tracking path to a terminal;

[0039] when the reward value is -100, feeding back information that the unmanned aerial vehicle tracking path needs to be adjusted, and the unmanned aerial vehicle automatically adjusts the unmanned aerial vehicle tracking path.

[0040] A wind power blade detection device based on unmanned aerial vehicle tracking, for executing the wind power blade detection method described above, characterized by comprising an initialization module, a target recognition module, a path generation module, and a defect detection module.

[0041] The initialization module is used to preset wind power blade detection initial information, which includes wind power blade initial information and unmanned aerial vehicle initial information, and preset an unmanned aerial vehicle working point. The unmanned aerial vehicle moves to the unmanned aerial vehicle working point based on the wind power blade detection initial information and acquires a working distance of the unmanned aerial vehicle.

[0042] The target recognition module is used for the unmanned aerial vehicle to perform wind power blade target recognition on the wind power blade through the wind power blade detection device and acquire a target wind power blade image. The wind power blade target recognition is realized through a YOLOv8 target detection algorithm.

[0043] The path generation module is used to generate an unmanned aerial vehicle tracking path based on the target wind power blade image through an improved path generation method.

[0044] The defect detection module is used for the unmanned aerial vehicle to perform wind power blade defect detection according to the unmanned aerial vehicle tracking path based on a PPO algorithm.

[0045] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the wind power blade detection method described above.

[0046] A storage medium containing computer executable instructions for performing the wind power blade detection method described above when executed by a computer processor.

[0047] The beneficial effects of the present application are:

[0048] (1) The wind power blade defect detection is performed by the unmanned aerial vehicle, the wind turbine is not required to be stopped, the efficiency of wind power generation is significantly improved, the impact of frequent start-stop operation on the service life of the wind power generator is reduced, intelligent detection is realized, the wind power blade is ensured to be comprehensively and real-timely detected, manpower is saved, the working efficiency of detection is improved, and the misjudgment risk caused by insufficient work experience or carelessness of the detection personnel is significantly reduced;

[0049] (2) The unmanned aerial vehicle tracking path is generated based on the target wind power blade boundary graph, the tracking route of the unmanned aerial vehicle is comprehensively designed, and defect detection is performed on all regions of the wind power blade;

[0050] (3) The wind power blade defect detection is performed based on the PPO algorithm according to the unmanned aerial vehicle tracking path, a complex mathematical model is not required to be established, and better generalization ability and robustness are achieved;

[0051] (4) The unmanned aerial vehicle tracking path adjustment information is output based on the reward value, the detection path of the unmanned aerial vehicle is adjusted in a timely manner, detection errors caused by unmanned aerial vehicle yaw are avoided, and serious consequences such as damage to the wind power blade are avoided;

[0052] (5) The working distance of the unmanned aerial vehicle is set, the industrial camera carried by the unmanned aerial vehicle is prevented from being too far from the wind power blade to cause unclear images and affect the success rate of wind power blade defect detection, and the industrial camera is also prevented from being too close to the wind power blade to cause image loss. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to facilitate the understanding of those skilled in the art, the present application will be further described below with reference to the accompanying drawings.

[0054] Figure 1 A wind power blade detection method based on unmanned aerial vehicle tracking according to the present application is shown in the flow chart. DETAILED DESCRIPTION

[0055] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purposes, the specific embodiments, structures, features and effects of the present application are described in detail below with reference to the accompanying drawings and preferred embodiments.

[0056] Working principle and use process of the present application:

[0057] Please refer to Figure 1 A wind power blade detection method based on unmanned aerial vehicle tracking, comprising:

[0058] S1: preset wind turbine blade detection initial information, the wind turbine blade detection initial information includes wind turbine blade initial information and unmanned aerial vehicle initial information;

[0059] S2: preset unmanned aerial vehicle working point, the unmanned aerial vehicle moves to the unmanned aerial vehicle working point based on the wind turbine blade detection initial information and obtains unmanned aerial vehicle working distance;

[0060] S3: the unmanned aerial vehicle performs wind turbine blade target identification on the wind turbine blade through a wind turbine blade detection device and obtains a target wind turbine blade image, and the wind turbine blade target identification is realized through a YOLOv8 target detection algorithm;

[0061] S4: generating an unmanned aerial vehicle tracking path based on the target wind turbine blade image through an improved path generation method;

[0062] S5: the unmanned aerial vehicle performs wind turbine blade defect detection based on a PPO algorithm according to the unmanned aerial vehicle tracking path.

[0063] In this embodiment, preset wind turbine blade detection initial information, the wind turbine blade detection initial information includes wind turbine blade initial information and unmanned aerial vehicle initial information, the wind turbine blade initial information specifically includes wind turbine blade rotating speed and wind turbine blade rotating center position, and the unmanned aerial vehicle initial information includes unmanned aerial vehicle initial position and detection initial distance, the unmanned aerial vehicle initial position is directly in front of the wind turbine blade rotating center position, and the detection initial distance is the distance between the wind turbine blade rotating center position and the unmanned aerial vehicle initial position.

[0064] In this embodiment, the unmanned aerial vehicle moves to the unmanned aerial vehicle working point based on the wind turbine blade detection initial information and obtains unmanned aerial vehicle working distance, the unmanned aerial vehicle working distance is the distance between the unmanned aerial vehicle working point and the wind turbine blade rotating center position, the unmanned aerial vehicle carries the wind turbine blade detection device, the wind turbine blade detection device includes a laser radar and an industrial camera, the laser radar is used to obtain the unmanned aerial vehicle working distance, ensures that the industrial camera carried by the unmanned aerial vehicle will not cause image unclear and affect the success rate of wind turbine blade defect detection due to being too far away from the wind turbine blade, and will not cause image loss due to being too close to the wind turbine blade.

[0065] In this embodiment, an unmanned aerial vehicle tracking path is generated based on the target wind turbine blade image through an improved path generation method, the tracking route of the unmanned aerial vehicle is comprehensively designed, the defect detection on all areas of the wind turbine blade is ensured, and the following steps can be specifically implemented:

[0066] S401: detecting the wind turbine blade edge of the target wind turbine blade image through a Canny edge detection algorithm and obtaining a target wind turbine blade boundary image;

[0067] S402: generating the UAV tracking path based on the target wind turbine blade boundary graph;

[0068] In this embodiment, the UAV tracking path is generated based on the target wind turbine blade boundary graph, which can be implemented by the following steps:

[0069] S402-1: extracting special boundary points of the target wind turbine blade boundary graph, the special boundary points being single wind turbine blade top points, the special boundary points including a first special boundary point, a second special boundary point, and a third special boundary point;

[0070] S402-2: connecting the special boundary points to obtain a special boundary point polygon, randomly selecting two edges of the special boundary point polygon to obtain a polygon center line, and obtaining a target wind turbine blade boundary graph barycenter through the intersection of the polygon center line;

[0071] S402-3: obtaining a wind turbine blade detection path set, the wind turbine blade detection path set including a first wind turbine blade detection path, a second wind turbine blade detection path, a third wind turbine blade detection path, and a fourth wind turbine blade detection path, the first wind turbine blade detection path being a directional path from the target wind turbine blade boundary graph barycenter to the first special boundary point, the second wind turbine blade detection path being a directional path from the first special boundary point to the second special boundary point, the third wind turbine blade detection path being a directional path from the second special boundary point to the target wind turbine blade boundary graph barycenter, and the fourth wind turbine blade detection path being a directional path from the target wind turbine blade boundary graph barycenter to the third special boundary point, and the UAV tracking path being obtained through the wind turbine blade detection path set.

[0072] In this embodiment, the UAV performs wind turbine blade defect detection based on the PPO algorithm according to the UAV tracking path, without the need to shut down the wind turbine generator, significantly improving the efficiency of wind power generation, reducing the impact of frequent start-stop operations on the service life of the wind turbine generator, and achieving intelligent detection to ensure comprehensive and real-time detection of wind turbine blades, saving manpower and improving the work efficiency of detection, while significantly reducing the risk of misjudgment caused by insufficient or careless work experience of the detection personnel. This can be implemented by the following steps:

[0073] S501: discretizing the UAV tracking path to obtain N UAV tracking path discrete points with equal distances;

[0074] S502: taking the target wind turbine blade boundary graph barycenter as a detection starting point, and traversing the UAV tracking path discrete points according to the PPO algorithm, without the need to establish a complex mathematical model, which has better generalization ability and robustness.

[0075] S503: The unmanned aerial vehicle detects defects of the wind turbine blade based on the Simple Net defect detection algorithm.

[0076] In this embodiment, the unmanned aerial vehicle takes the gravity center of the target wind turbine blade boundary graph as the detection starting point, and traverses the discrete points of the unmanned aerial vehicle tracking path according to the PPO algorithm. Specifically, the following steps can be implemented:

[0077] S502-1: A preset shooting period is set, and a set of wind turbine blade images is obtained by the industrial camera according to the shooting period;

[0078] S502-2: The center point of the wind turbine blade image is obtained, a discrete point distance threshold is preset, and a target distance is obtained, the target distance being the distance between the center point of the wind turbine blade image and the discrete point of the unmanned aerial vehicle tracking path. When the target distance is less than or equal to the discrete point distance threshold, the discrete point of the unmanned aerial vehicle tracking path is set as the next target point, which is the discrete point of the unmanned aerial vehicle tracking path reached by the unmanned aerial vehicle in the next step;

[0079] S502-3: The discrete points of the unmanned aerial vehicle tracking path are traversed based on the PPO algorithm, and the minimum discrete distance is obtained, which is the distance between the unmanned aerial vehicle and the nearest discrete point of the unmanned aerial vehicle tracking path. The algorithm input state is obtained, which is the input state of the PPO algorithm at time t. The algorithm input state includes the working distance difference, the target distance, the discrete point direction, the discrete point distance, and the wind turbine speed. The working distance difference is the difference between the working distance of the unmanned aerial vehicle and the ideal working distance of the unmanned aerial vehicle. The discrete point direction is the direction of the discrete point of the unmanned aerial vehicle tracking path relative to the center point of the wind turbine blade image in the wind turbine blade image. The discrete point distance is the distance between the discrete point of the unmanned aerial vehicle tracking path and the gravity center of the target wind turbine blade boundary graph. The expression of the algorithm input state is:

[0080] s t =(Dis t ,dis t ,θ t ,Δ t ,ω)

[0081] Dis t =min(Radar t )-ΔD,

[0082]

[0083] where s t is the algorithm input state, Dist is the working distance difference, dis t is the target distance, θ t is the discrete point orientation, Δ t is the discrete point distance, ω is the wind turbine blade speed, min(Radar t ) is the minimum discrete distance, ΔD is the UAV working distance, x j is the special boundary point horizontal coordinate, y j is the special boundary point vertical coordinate, x i is the UAV tracking path discrete point horizontal coordinate, y i is the UAV tracking path discrete point vertical coordinate, x0 is the target wind turbine blade boundary graph barycenter horizontal coordinate, y0 is the target wind turbine blade boundary graph barycenter vertical coordinate;

[0084] S502-4: Obtain the reward value of the PPO algorithm, output UAV tracking path adjustment information based on the reward value, and timely adjust the detection path of the UAV to avoid detection errors caused by UAV yaw and even serious consequences of wind turbine blade damage, the UAV tracking path adjustment information includes UAV tracking path no adjustment information, UAV tracking path abnormal information, and UAV tracking path adjustment information, and the expression of the reward value is:

[0085]

[0086] Wherein, rt is the reward value, Dis t is the working distance difference, ε is the discrete point distance threshold, Tanh is the hyperbolic tangent, dis t is the target distance, θ t is the discrete point orientation, min(Radar t ) is the minimum discrete distance, ΔD is the UAV working distance;

[0087] Determine the reward value, when the reward value is 100, feedback the UAV tracking path no adjustment information;

[0088] When the reward value is in the open interval (-100, 100), feedback the UAV tracking path abnormal information to a terminal;

[0089] When the reward value is -100, feedback the UAV tracking path adjustment information, and the UAV automatically adjusts the UAV tracking path;

[0090] A wind turbine blade detection device based on UAV tracking, comprising an initialization module, a target recognition module, a path generation module, and a defect detection module.

[0091] The initialization module is used for presetting wind power blade detection initial information, the wind power blade detection initial information including wind power blade initial information and unmanned aerial vehicle initial information, presetting an unmanned aerial vehicle working point, and the unmanned aerial vehicle moving to the unmanned aerial vehicle working point and acquiring an unmanned aerial vehicle working distance based on the wind power blade detection initial information;

[0092] The target identification module is used for the unmanned aerial vehicle to perform wind power blade target identification on the wind power blade through a wind power blade detection device and acquire a target wind power blade image, and the wind power blade target identification is realized through a YOLOv8 target detection algorithm.

[0093] The path generation module is used for generating an unmanned aerial vehicle tracking path through an improved path generation method based on the target wind power blade image.

[0094] The defect detection module is used for the unmanned aerial vehicle to perform wind power blade defect detection according to the unmanned aerial vehicle tracking path based on a PPO algorithm.

[0095] The computer storage medium of the embodiment of the application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection with one or more conductive wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or component.

[0096] The computer readable signal medium can include a data signal propagated in a baseband or as a part of a carrier wave, in which a computer readable program code is carried. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, device or component.

[0097] The computer readable media on which the program code can be carried by any suitable medium, including but not limited to wireless, wired, optical fiber cable, RF, and the like, or any suitable combination of these. Computer program code for carrying out operations of the present application can be written in one or more programming languages, or combinations of languages, including object oriented, such as Java, Smalltalk, C++, and conventional procedural, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0098] The above description is only the preferred embodiment of the present application, and is not intended to limit the present application in any form. Although the present application has been disclosed as above with the preferred embodiment, it is not intended to limit the present application, and any person skilled in the art can make some changes or modifications to the equivalent embodiments within the scope of the technical solution of the present application, without departing from the technical solution of the present application. Any simple modification, equivalent change and modification of the above embodiments, which does not depart from the technical solution of the present application, and is based on the technical essence of the present application, still belongs to the scope of the technical solution of the present application.

Claims

1. A wind turbine blade detection method based on unmanned aerial vehicle tracking, characterized in that, The implementation of the wind power blade detection method comprises the following steps: S1: presetting wind power blade detection initial information, the wind power blade detection initial information comprising wind power blade initial information and unmanned aerial vehicle initial information, the wind power blade initial information specifically comprising wind power blade rotating speed and wind power blade rotating center position, the unmanned aerial vehicle initial information comprising unmanned aerial vehicle initial position and detection initial distance, the unmanned aerial vehicle initial position being directly in front of the wind power blade rotating center position, and the detection initial distance being the distance between the wind power blade rotating center position and the unmanned aerial vehicle initial position; S2: presetting an unmanned aerial vehicle working point, the unmanned aerial vehicle moving to the unmanned aerial vehicle working point based on the wind power blade detection initial information and acquiring unmanned aerial vehicle working distance, the unmanned aerial vehicle working distance being the distance between the unmanned aerial vehicle working point and the wind power blade rotating center position, the unmanned aerial vehicle carrying a wind power blade detection device, the wind power blade detection device comprising a laser radar and an industrial camera, and the laser radar being used to acquire the unmanned aerial vehicle working distance; S3: the unmanned aerial vehicle performing wind power blade target identification on the wind power blade through the wind power blade detection device and acquiring a target wind power blade image, the wind power blade target identification being achieved through a YOLOv8 target detection algorithm; S4: generating an unmanned aerial vehicle tracking path based on the target wind power blade image through an improved path generation method; S401: detecting the wind power blade edge of the target wind power blade image through a Canny edge detection algorithm and acquiring a target wind power blade boundary image; S402: generating the unmanned aerial vehicle tracking path based on the target wind power blade boundary image; S5: the unmanned aerial vehicle performing wind power blade defect detection according to the unmanned aerial vehicle tracking path based on a PPO algorithm; S501: discretizing the unmanned aerial vehicle tracking path to acquire N unmanned aerial vehicle tracking path discrete points with equal distances; S502: the unmanned aerial vehicle taking the gravity center of the target wind power blade boundary image as a detection starting point and traversing the unmanned aerial vehicle tracking path discrete points according to a PPO algorithm; S503: the unmanned aerial vehicle performing the wind power blade defect detection based on a Simple Net defect detection algorithm.

2. The wind turbine blade inspection method of claim 1, wherein, The step S402 specifically comprises: S402-1: extracting special boundary points of the target wind power blade boundary image, the special boundary points being single wind power blade top end points, and the special boundary points comprising a first special boundary point, a second special boundary point and a third special boundary point; S402-2: connecting the special boundary points to acquire a special boundary point polygon, randomly selecting two edges of the special boundary point polygon to make a center line to acquire a polygon center line, and acquiring a target wind power blade boundary image gravity center through the intersection of the polygon center lines; S402-3: acquiring a wind power blade detection path set and acquiring the unmanned aerial vehicle tracking path based on the wind power blade detection path set.

3. The wind turbine blade inspection method of claim 2, wherein, The step S402-3 is specifically: The wind power blade detection path set includes a first wind power blade detection path, a second wind power blade detection path, a third wind power blade detection path, and a fourth wind power blade detection path. The first wind power blade detection path is a directional path connected from the target wind power blade boundary graph center to the first special boundary point. The second wind power blade detection path is a directional path connected from the first special boundary point to the second special boundary point. The third wind power blade detection path is a directional path connected from the second special boundary point to the target wind power blade boundary graph center. The fourth wind power blade detection path is a directional path connected from the target wind power blade boundary graph center to the third special boundary point. The UAV tracking path is obtained through the wind power blade detection path set.

4. The wind turbine blade inspection method of claim 1, wherein, The step S502 specifically includes: S502-1: A shooting cycle is preset, and a wind power blade image set is obtained through the industrial camera according to the shooting cycle. The wind power blade image set is a set of wind power blade images. S502-2: A wind power blade image center point of the wind power blade image is obtained, a discrete point distance threshold value is preset, and a target distance is obtained. The target distance is the distance between the wind power blade image center point and the UAV tracking path discrete point. When the target distance is less than or equal to the discrete point distance threshold value, the UAV tracking path discrete point is set as a next target point. The next target point is the UAV tracking path discrete point reached by the UAV in the next step. S502-3: The UAV tracking path discrete point is traversed based on the PPO algorithm, and a minimum discrete distance is obtained. The minimum discrete distance is the distance between the UAV and the nearest UAV tracking path discrete point. An algorithm input state is obtained. The algorithm input state is the input state of the PPO algorithm at time t. The algorithm input state includes a working distance difference value, the target distance, a discrete point direction, a discrete point distance, and the wind power blade rotating speed. The working distance difference value is the difference between the UAV working distance and the ideal UAV working distance. The discrete point direction is the direction of the UAV tracking path discrete point relative to the wind power blade image center point of the wind power blade image. The discrete point distance is the distance between the UAV tracking path discrete point and the target wind power blade boundary graph center. The expression of the algorithm input state is: ， ， ， ， ， wherein s t is an algorithm input state, Dis t is a working distance difference value, dis t is a target distance, is a discrete point azimuth, is a discrete point distance, is a wind turbine blade rotating speed, is a minimum discrete distance, is a UAV working distance, x j is a special boundary point abscissa, y j is a special boundary point ordinate, x i is a UAV tracking path discrete point abscissa, y i is a UAV tracking path discrete point ordinate, x0is a target wind turbine boundary figure barycenter abscissa, y0is a target wind turbine boundary figure barycenter ordinate; S502-4: A reward value of the PPO algorithm is obtained, and UAV tracking path adjustment information is output based on the reward value. The UAV tracking path adjustment information includes UAV tracking path no adjustment information, UAV tracking path abnormal information, and UAV tracking path adjustment information.

5. The wind turbine blade inspection method of claim 4, wherein, The step S502-4 specifically includes: The expression of the reward value is: ， Wherein, rt is a reward value, Dis t is a working distance difference value, is a discrete point distance threshold value, Tanh is a hyperbolic tangent, dis t is a target distance, is a discrete point orientation, is a minimum discrete distance, is a working distance of the unmanned aerial vehicle; The reward value is determined. When the reward value is 100, the UAV tracking path no adjustment information is fed back. When the reward value is in the open interval (-100, 100), the UAV tracking path abnormal information is fed back to a terminal. When the reward value is -100, feedback is given that the UAV tracking path needs to be adjusted, and the UAV automatically adjusts the UAV tracking path.

6. A wind turbine blade inspection apparatus based on drone tracking, characterized in that, The method comprises an initialization module, a target identification module, a path generation module, and a defect detection module. The initialization module is configured to preset wind turbine blade detection initial information, which comprises wind turbine blade initial information and UAV initial information, and preset a UAV working point. The target identification module is configured to perform wind turbine blade target identification on a wind turbine blade by the wind turbine blade detection device and obtain a target wind turbine blade image. The path generation module is configured to generate a UAV tracking path based on the target wind turbine blade image by an improved path generation method. The path generation module further comprises detecting a wind turbine blade edge of the target wind turbine blade image based on a Canny edge detection algorithm and obtaining a target wind turbine blade boundary image, and generating the UAV tracking path based on the target wind turbine blade boundary image. The defect detection module is configured to perform wind turbine blade defect detection based on a PPO algorithm according to the UAV tracking path.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the wind turbine blade detection method of any one of claims 1-5.

8. A storage medium containing computer-executable instructions, characterized in that, The computer executable instructions, when executed by the computer processor, are configured to perform the wind turbine blade detection method of any one of claims 1-5.

Images