A UAV system for ground inspection and fault location of solar photovoltaic power stations

Abstract

The present invention relates to the field of inspection by patrol drones, and in particular to a drone system for ground inspection and fault location of solar photovoltaic power stations; technical solution: a drone system for ground inspection and fault location of solar photovoltaic power stations, comprising a patrol drone, a patrol drone carrying platform, a patrol drone lifting platform, a communication module, a drone tracking module, a background server end and an intermediate processing end; the present invention carries a drone tracking module on the patrol drone carrying platform, thereby eliminating the need to install image acquisition equipment and multiple sensors for positioning on the patrol drone, thereby reducing the structural complexity of the patrol drone, thereby reducing the requirements for the patrol drone, thereby increasing the endurance of the patrol drone, and reducing the hardware requirements of the system, and because the background image of the positioning based image collected in this way is the sky, the requirements for the environment are relatively low, and fuzzy positioning can still be achieved in a working environment with low visibility.

Description

Technical Field

[0001] The present invention relates to the field of inspection by unmanned aerial vehicles (UAVs), and in particular to an unmanned aerial vehicle (UAV) system for ground inspection and fault location of solar photovoltaic power stations. Background Art

[0002] The solar photovoltaic power station fault inspection drone system is a technical solution specially designed for automated inspection and monitoring of solar photovoltaic power stations. By introducing the drone system, solar photovoltaic power stations can greatly improve the efficiency and accuracy of inspections while reducing the cost and risk of manual inspections.

[0003] In the prior art, the intelligent control systems of inspection drones mainly include control systems based on machine vision navigation and control systems based on satellite positioning. The control system based on machine vision navigation needs to be equipped with a variety of hardware such as a variety of acquisition devices, processors, and high-precision equipment on the inspection drone, and has high performance requirements for the inspection drone, thereby increasing the hardware cost of the system, and will increase the weight of the inspection drone and the energy consumption of the inspection drone, thereby reducing the endurance of the inspection drone. Due to the performance requirements of the inspection drone, the cameras and sensors on the inspection drone have a performance upper limit, and when the ground image is relatively simple or too complex, the inspection drone is prone to positioning errors or even failure to locate. On the one hand, the control system based on satellite positioning has high requirements on communication quality, which can easily lead to signal loss, and has certain limitations on accuracy and communication real-time performance, which is not reliable for the fully automated control of the inspection drone. Summary of the invention

[0004] In order to overcome the control system based on machine vision navigation and the control system based on satellite positioning, the control system based on machine vision navigation needs to carry a variety of hardware such as a variety of acquisition devices, processors and high-precision equipment on the inspection drone, which has high performance requirements for the inspection drone, thereby increasing the hardware cost of the system, and will increase the weight of the inspection drone and the energy consumption of the inspection drone, thereby reducing the endurance of the inspection drone. Due to the performance requirements of the inspection drone, the cameras and sensors on the inspection drone have a performance ceiling, and when the ground image is relatively simple or too complex, the inspection drone is prone to positioning errors or even failure to locate. On the one hand, the control system based on satellite positioning has high requirements for communication quality, which can easily lead to signal loss, and has certain limitations on accuracy and communication real-time performance, which is not a reliable problem for the fully automated control of the inspection drone.

[0005] The technical solution of the present invention is: an unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations, comprising:

[0006] A patrol drone is used to carry image acquisition equipment to inspect the solar photovoltaic power station using the image acquisition equipment, wherein the image acquisition equipment carried includes an infrared image acquisition equipment and a visible light image acquisition equipment;

[0007] The inspection drone carrying platform is used to carry the inspection drone and provide charging and transportation services for the inspection drone. The inspection drone carrying platform includes:

[0008] The inspection drone lifting platform is used to provide a take-off and landing platform for the inspection drone, and the inspection drone lifting platform is provided with a mark for identification and positioning;

[0009] The communication module is used to build a communication network to achieve communication between the inspection drone and the inspection drone carrying platform;

[0010] The drone tracking module includes a tracking camera, which is used to track and collect images of the patrol drone while it is working, and to control the patrol drone;

[0011] The data processing end is used to process and identify the image data collected by the image acquisition device to realize the fault identification and fault location of the solar photovoltaic power station. The data processing end includes:

[0012] The backend server is used to process and archive the data collected by the image acquisition equipment, and to control the inspection drone carrying platform and patrol drones;

[0013] The intermediate processing end is used to process and identify the data collected by the drone tracking module and control the drone tracking module, wherein the intermediate processing end is arranged on the inspection drone carrying platform.

[0014] Preferably, an environmental sensing module is also provided on the inspection UAV carrying platform, and the environmental sensing module is used to detect whether the environment where the inspection UAV carrying platform is located is suitable for the inspection UAV to take off. The environmental sensing module includes a wind speed sensor, a light intensity sensor, a rain sensor and a visibility monitor. The wind speed sensor is used to detect the wind speed parameters of the environment where the inspection UAV carrying platform is located, the light intensity sensor is used to detect the light intensity parameters of the environment where the inspection UAV carrying platform is located, the rain sensor is used to detect the light intensity parameters of the environment where the inspection UAV carrying platform is located, and the visibility monitor is used to detect the visibility parameters of the environment where the inspection UAV carrying platform is located.

[0015] Preferably, when the background server processes the data collected by the image acquisition device, the following steps are included:

[0016] S11: extracting and preprocessing a plurality of groups of images collected by the visible light image acquisition device, wherein the preprocessing method adopted includes image enhancement processing based on the spatial domain and image filtering processing based on the Laplace filtering method;

[0017] S12: stitching the multiple groups of images collected by the visible light image acquisition device into a total device image, wherein the total device image is one or more groups of images having images of all devices in the solar photovoltaic power station;

[0018] S13: splicing several groups of infrared images collected by the infrared image acquisition device based on the processing results of steps S11 and S12 to obtain an overall infrared image of the photovoltaic power station equipment;

[0019] S14: Process and identify the equipment visible light total image and the equipment infrared total image to identify equipment faults and defects in the solar photovoltaic power station.

[0020] Preferably, when the backend server stitches a plurality of groups of images acquired by the visible light image acquisition device, the backend server comprises the following steps:

[0021] S1201: performing distortion correction on a plurality of groups of images collected by a visible light image collection device to remove image distortion caused by a camera lens structure;

[0022] S1202: extracting sift feature points from several groups of visible light images, matching the sift feature points of two adjacent visible light images, and using the RANSAC algorithm to screen feature point pairs after matching, and eliminating erroneous feature point pairs in the feature point pairs;

[0023] S1203: Use the DLT algorithm to estimate the perspective change matrix of the remaining feature point pairs, and then use the SeamFinding algorithm to find the best stitching seam, and find the best stitching seam of the images collected by two adjacent groups of high-definition cameras; finally, use the Laplacian pyramid to decompose the original image into sub-images of different scales by differentiating the Gaussian pyramids of two adjacent layers, and perform weighted averaging on each sub-image to obtain the fusion result of each layer, and finally perform reverse reconstruction of the pyramid to obtain the final fusion effect image, while retaining the covered overlapping parts;

[0024] S1204: performing image detection on the final fusion effect image, and using an image detection algorithm based on the HSV color space model to identify illumination interference in the final fusion effect image;

[0025] S1205: Eliminate the illumination interference in the final fusion effect image, and detect the overlapping images of the eliminated part. If there is no illumination interference in the overlapping images, fill the overlapping images into the eliminated part to obtain the total visible light image of the device.

[0026] As a preferred embodiment, the backend server side includes the following steps when controlling the inspection drone carrying platform and the patrol drone:

[0027] S21: According to the inspection area of ​​the solar photovoltaic power station, the inspection drone carrying platform is planned to determine the driving path of the inspection drone carrying platform;

[0028] S22: setting flight parameters of the inspection drone according to the various environmental parameters collected by the environmental sensing module, wherein the flight parameters of the inspection drone include flight speed, flight altitude and endurance distance;

[0029] S23: planning the inspection path of the inspection drone according to the set flight parameters of the inspection drone;

[0030] S24: According to the inspection path planning of the inspection drone, the path planning of the inspection drone carrying platform is modified.

[0031] Preferably, when the background server sets the flight parameters of the inspection drone according to the various environmental parameters collected by the environmental sensing module, the following steps are included:

[0032] S2201: Cleaning the data collected by various sensors in the environment monitoring module;

[0033] S2202: normalize the cleaned data, wherein the data is scaled to the interval [0,1] by using maximum and minimum normalization, wherein the principle formula is:

[0034]

[0035] in, represents the normalized data, represents the raw data samples of the sensor, Represents a manually set dataset The minimum value of Represents a manually set dataset The maximum value of

[0036] S2203: Analyze the data after data cleaning and data preprocessing to obtain the environmental interference coefficient in the solar photovoltaic power station, wherein the principle formula for analyzing the data after data cleaning and data preprocessing is:

[0037] ;

[0038] in, is the environmental interference coefficient obtained by analysis, is the value of the nth sensor after maximum and minimum normalization processing, For The nonlinear adjustment function of is a constant coefficient, where the nonlinear adjustment function Nonlinear adjustment function obtained by summarizing historical data;

[0039] S2204: Setting the flight parameters of the inspection UAV according to the environmental interference coefficient.

[0040] Preferably, the drone tracking module includes the following steps when tracking and collecting images of the patrol drone when it is working and controlling the patrol drone:

[0041] S31: After the inspection drone takes off, start the tracking camera to collect images of the inspection drone, and use a video-based image tracking algorithm to identify and track the images of the inspection drone;

[0042] S32: Obtaining the location of the inspection drone according to the angle of the tracking camera, the location of the tracking camera, and the image collected by the tracking camera;

[0043] S33: Control the inspection drone according to the positioning of the inspection drone, so that the inspection drone performs inspection along the set path.

[0044] Preferably, a modular cleaning device is also provided on the inspection drone carrying platform, and the modular cleaning device is used to clean pollutants on the solar photovoltaic panels, and a mounting structure corresponding to the modular cleaning device is provided at the bottom of the inspection drone.

[0045] Preferably, the modular cleaning equipment includes a mounting bracket, a drive motor, a connecting component, a brushing component and a flushing component. The mounting bracket is configured as a cylindrical structure, the drive motor is arranged on the top surface of the mounting bracket, the connecting component is arranged inside the mounting bracket, the brushing component is arranged inside the connecting component, and the flushing component is arranged inside the connecting component. A carrying component is also arranged on the inspection drone carrying platform, and the carrying component is used to fix and maintain the modular cleaning equipment.

[0046] Preferably, the connecting assembly includes an output shaft, a connecting sleeve, a connecting shaft, a buffer spring and a mounting shell, the output shaft is arranged inside the mounting bracket, one end of the output shaft is fixedly connected to the output end of the driving motor, the other end of the output shaft is fixedly connected to the connecting sleeve, the buffer spring is arranged inside the connecting sleeve, one end of the buffer spring is fixedly connected to the internal top surface of the connecting sleeve, the other end of the buffer spring is fixedly connected to the connecting shaft, the connecting shaft is slidably connected to the connecting sleeve, and the mounting shell is arranged at the bottom end of the connecting shaft.

[0047] Preferably, the brushing assembly includes a mounting plate and a fixed spring, the mounting plate is arranged below the mounting shell, the fixed springs are arranged in multiple groups, one end of the fixed spring is fixedly connected to the bottom surface of the mounting plate, the other end of the fixed spring is fixedly connected to the inner top surface of the mounting shell, and bristles are arranged on the bottom surface of the mounting plate.

[0048] Preferably, the flushing assembly includes a storage bag, a connecting pipe, a one-way valve and a nozzle. The storage bag is arranged inside the mounting shell, the connecting pipe is arranged at the center of the mounting plate, one end of the connecting pipe is through-connected with the storage bag, the one-way valve is arranged inside the connecting pipe, the nozzle is arranged in multiple groups, the nozzle is arranged on the mounting plate, and the nozzle is through-connected with the storage bag.

[0049] Preferably, the carrying assembly includes a mounting platform, a storage box, a conveying pipeline and a pressure pump. The mounting platform is arranged on the inspection UAV carrying platform. A truncated cone-shaped groove is provided on the top surface of the mounting platform. The storage box is arranged below the mounting platform. The conveying pipeline is arranged on the inner bottom surface of the truncated cone-shaped groove. The conveying pipeline is connected with the storage box, and the pressure pump is arranged on the conveying pipeline.

[0050] Preferably, the modular cleaning equipment also includes an auxiliary component, which includes a mounting plate, an auxiliary motor, a driving gear, a mounting shaft, a transmission gear and a rotating arm. The mounting plate is arranged above the mounting bracket, and the mounting plate is fixedly connected to the mounting bracket. The auxiliary motor is arranged on the bottom surface of the mounting plate, and the driving gear is arranged on the output end of the auxiliary motor. The mounting shaft is arranged in two groups, and the mounting shaft is rotatably connected to the mounting plate. The transmission gear is arranged in two groups, and the transmission gear is arranged on the mounting shaft, and the transmission gear is meshed with the driving gear. The rotating arm is arranged in two groups, and the rotating arm is arranged in a semicircular structure, and the rotating arm is fixedly connected to the mounting shaft.

[0051] Preferably, an arc-shaped groove is provided on the top surface of the mounting disk, and a magnetic block is arranged at the center of the top surface of the mounting disk.

[0052] Preferably, the mounting structure includes a connecting rope and a hook, one end of the connecting rope is fixedly connected to the bottom end of the inspection drone, and the other end of the connecting rope is fixedly connected to the hook.

[0053] Preferably, after the backend server processes and identifies the total visible light image and the total infrared image of the device and identifies the equipment faults and defects in the solar photovoltaic power station, the following steps are included:

[0054] S31: classifying equipment failures and defects in the solar photovoltaic power station and determining the types of equipment failures and defects;

[0055] S32: If the equipment failure and defect type is contaminants on the photovoltaic panels, the type of contaminants is identified, and different contaminant cleaning solutions are retrieved, and the retrieved contaminant cleaning solutions are sent to the intermediate processing end;

[0056] S33: The intermediate processing end controls the pressure pump according to the received cleaning plan to deliver different amounts of cleaning liquid into the storage capsule, and controls the inspection drone to connect with the auxiliary component through the mounting structure after the inspection drone completes the staged inspection target;

[0057] S34: Control the inspection drone to fly above the pollutants and descend, place the modular cleaning device above the pollutants, and clean the pollutants according to the set rotation speed and cleaning time.

[0058] Beneficial effects of the present invention:

[0059] 1. By installing a drone tracking module on the inspection drone platform, it is unnecessary to install image acquisition equipment and multiple sensors for positioning on the inspection drone, thereby reducing the structural complexity of the inspection drone, thereby reducing the requirements for the inspection drone, thereby increasing the endurance of the inspection drone and reducing the hardware requirements of the system. In addition, since the background image of the positioning image collected by this method is the sky, the requirements for the environment are relatively low, and fuzzy positioning can still be achieved in a working environment with low visibility;

[0060] 2. When using the image acquisition equipment carried by the inspection drone to collect images of the photovoltaic power station, since the solar photovoltaic panels are relatively smooth planes, light interference such as sun reflection spots will be formed on the solar photovoltaic panels. This light interference will also affect the recognition of infrared images, which may cover up the hot spots in the illuminated area, causing missed detection of hot spots, or be regarded as rectangular hot spots, causing false detection of hot spots. Therefore, when stitching images, images without light interference of the same object are found, and the real scenes found are used as references to reconstruct the real scenes in the light interference area, so as to achieve the effect of eliminating light interference, thereby improving the performance of image recognition and increasing the accuracy of fault detection and fault identification;

[0061] 3. The flight parameters of the inspection drone are adjusted through the data collected by the environmental sensing module to ensure the stability of the inspection drone. The accuracy of the inspection can be ensured by controlling the clarity and recognizability of the image. According to the flight parameters of the inspection drone, and due to the limitations of the image acquisition equipment, the effective acquisition range of the image acquisition equipment is different when the inspection drone flies at different altitudes. Therefore, according to the different flight parameters of the inspection drone, the inspection path of the inspection drone is replanned to ensure the effectiveness and comprehensiveness of the inspection image acquisition. At the same time, the inspection path of the inspection drone carrying platform is replanned, such as changing the temporary stop point and driving route for the inspection, which can prevent the inspection drone from having insufficient endurance and ensure the efficiency of the inspection.

[0062] 4. Modular cleaning equipment can be used to remove pollutants detected by drone inspections, thereby preventing pollutants from affecting the power generation efficiency of solar panels and preventing the presence of pollutants from causing hot spot effects on solar panels and causing damage to them. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 Shown is a schematic diagram of the structure of the drone system for ground inspection and fault location of solar photovoltaic power stations of the present invention;

[0064] Figure 2 Shown is a schematic diagram of part of the workflow of the background processing end of the drone system for ground inspection and fault location of solar photovoltaic power stations of the present invention;

[0065] Figure 3 Shown is a schematic diagram of part of the workflow of the background processing end of the drone system for ground inspection and fault location of solar photovoltaic power stations of the present invention;

[0066] Figure 4 Shown is a schematic diagram of the three-dimensional structure of a modular cleaning device in a drone system for ground inspection and fault location of a solar photovoltaic power station according to the present invention;

[0067] Figure 5 Shown is a schematic diagram of a cross-sectional three-dimensional structure of a modular cleaning device in a drone system for ground inspection and fault location of a solar photovoltaic power station according to the present invention;

[0068] Figure 6 Another cross-sectional three-dimensional structural schematic diagram of a modular cleaning device in a drone system for ground inspection and fault location of a solar photovoltaic power station according to the present invention is shown;

[0069] Figure 7 Shown is a three-dimensional structural schematic diagram of the components carried in the drone system for ground inspection and fault location of solar photovoltaic power stations of the present invention;

[0070] Figure 8 Shown is a three-dimensional structural schematic diagram of the mounting structure of the drone system for ground inspection and fault location of solar photovoltaic power stations of the present invention;

[0071] Description of reference numerals: 1. mounting bracket; 2. driving motor;

[0072] 3. Connecting assembly; output shaft; 302. Connecting sleeve; 303. Connecting shaft; 304. Buffer spring; 305. Mounting housing;

[0073] 4. Brushing assembly; 401. Mounting plate; 402. Fixing spring;

[0074] 5. Flushing assembly; 501. Storage bag; 502. Connecting pipe; 503. One-way valve; 504. Spray nozzle;

[0075] 6. Carrying components; 601. Mounting platform; 602. Storage box; 603. Delivery pipeline; 604. Pressure pump;

[0076] 7. Auxiliary components; 701. Mounting plate; 702. Auxiliary motor; 703. Driving gear; 704. Mounting shaft; 705. Transmission gear; 706. Rotating arm;

[0077] 801. Connecting rope; 802. Hook. DETAILED DESCRIPTION

[0078] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0079] Embodiment 1

[0080] See also Figure 1 The present invention provides an embodiment: an unmanned aerial vehicle system for ground inspection and fault location of a solar photovoltaic power station, comprising:

[0081] A patrol drone is used to carry image acquisition equipment to inspect the solar photovoltaic power station using the image acquisition equipment, wherein the image acquisition equipment carried includes an infrared image acquisition equipment and a visible light image acquisition equipment;

[0082] The inspection drone carrying platform is used to carry the inspection drone and provide charging and transportation services for the inspection drone. It uses an intelligent unmanned vehicle or an intelligent robot as the basic platform and is equipped with a lifting platform for the inspection drone, a communication module and a drone tracking module. The inspection drone takes off and lands on the lifting platform for the inspection drone. At the same time, the inspection drone carrying platform is equipped with an intermediate processing end. The inspection drone communicates with the intermediate processing end through the communication module. The communication module uses WiFi technology, eMTC technology, 4G technology, 5G technology, LoRa technology or other long-distance communication technologies to build a communication network. The choice of communication module is determined according to the needs of the solar photovoltaic power station. At the same time, the inspection drone carrying platform is also equipped with a drone tracking module, that is, a tracking camera for tracking and collecting images of the inspection drone, which is used to collect images when the inspection drone is working, and use the intermediate processing end to identify and analyze the images, so as to determine the coordinates of the inspection drone.

[0083] The data processing end includes a background server end for analyzing the images collected by the image acquisition device carried by the patrol drone and an intermediate processing end for real-time control of the patrol drone. The background server end is installed in the computer room server of the solar photovoltaic power station, and performs overall control of the patrol drone carrying platform and the patrol drone, such as patrol plan generation, patrol plan adjustment and solar photovoltaic power station fault analysis. The background server end is used to process and analyze the images collected by the tracking camera, identify the image and relative position of the patrol drone, and thus accurately control the patrol drone;

[0084] Specifically, the intelligent control systems of inspection drones in the prior art mainly include control systems based on machine vision navigation and control systems based on satellite positioning. The control systems based on machine vision navigation need to be equipped with a variety of hardware such as a variety of acquisition devices, processors, and high-precision equipment on the inspection drones, which has high performance requirements for the inspection drones, thereby increasing the hardware cost of the system, and increasing the weight and energy consumption of the inspection drones, thereby reducing the endurance of the inspection drones. In addition, due to the performance requirements of the inspection drones, the cameras and sensors on the inspection drones have a performance upper limit, and when the ground image is relatively simple or too complex, the inspection drones are prone to positioning errors or even failure to locate. On the one hand, the control system based on satellite positioning has high requirements on communication quality, which can easily lead to signal loss, and has certain limitations on accuracy and communication real-time performance. It is not reliable for the fully automated control of inspection drones. By equipping the inspection drone with a drone tracking module on the inspection drone carrying platform, it is unnecessary to install image acquisition equipment and various sensors for positioning on the inspection drone, thereby reducing the structural complexity of the inspection drone, thereby reducing the requirements for the inspection drone, thereby increasing the endurance of the inspection drone, and reducing the hardware requirements of the system. In addition, since the background image of the positioning based image collected by this method is the sky, the requirements for the environment are relatively low, and fuzzy positioning can still be achieved in a working environment with low visibility.

[0085] Furthermore, an environmental sensing module is also provided on the inspection UAV carrying platform. The environmental sensing module is used to detect whether the environment where the inspection UAV carrying platform is located is suitable for the inspection UAV to take off. The environmental sensing module includes a wind speed sensor, a light intensity sensor, a rain sensor and a visibility monitor. The wind speed sensor is used to detect the wind speed parameters of the environment where the inspection UAV carrying platform is located. The light intensity sensor is used to detect the light intensity parameters of the environment where the inspection UAV carrying platform is located. The rain sensor is used to detect the light intensity parameters of the environment where the inspection UAV carrying platform is located. The visibility monitor is used to detect the visibility parameters of the environment where the inspection UAV carrying platform is located.

[0086] Furthermore, the drone tracking module includes the following steps when tracking and collecting images of the patrol drone while it is working and controlling the patrol drone:

[0087] S31: After the inspection drone takes off, start the tracking camera to collect images of the inspection drone, and use a video-based image tracking algorithm to identify and track the images of the inspection drone;

[0088] S32: Obtaining the location of the inspection drone according to the angle of the tracking camera, the location of the tracking camera, and the image collected by the tracking camera;

[0089] S33: Control the inspection drone according to the positioning of the inspection drone, so that the inspection drone performs inspection along the set path.

[0090] Specifically, when using a video-based image tracking algorithm to identify and track the image of the inspection drone, an image recognition algorithm based on Fast R-CNN is used to identify the image of the inspection drone, and a MEDIANFLOW tracker is used to track the image of the inspection drone. The horizontal and vertical angles of the tracking camera are adjusted according to the position change of the image of the inspection drone, so that the image of the inspection drone always appears in the center of the image of the tracking camera.

[0091] The MEDIANFLOW tracker can track the forward and backward directions of an object in real time and measure the difference between these two tracks, minimizing this forward-backward error, allowing them to reliably detect tracking failures and select reliable tracks in video sequences. In testing, it was found that this tracker works best when the motion is predictable and the objects are small.

[0092] Furthermore, when obtaining the location of the inspection drone based on the angle of the tracking camera, the location of the tracking camera, and the image collected by the tracking camera, the calculation principle is:

[0093] S3201: using the recognition result of step S31, segmenting the image of the inspection drone to obtain the image of the inspection drone;

[0094] S3202: Measure and calculate the distance between the two propeller brackets of the inspection drone in the image ;

[0095] S3203: Calculate the distance between the inspection drone and the tracking camera, where the calculation principle formula is:

[0096] ;

[0097] Where d is the distance between the inspection drone and the tracking camera. is the actual distance between the two propeller brackets;

[0098] S3303: determining the angle between the inspection drone and the tracking camera according to the position of the inspection drone in the image and the angle of the tracking camera;

[0099] S3304: According to the positioning of the platform carrying the inspection drone, the positioning of the tracking camera is obtained, and the positioning of the inspection drone is calculated according to the positioning of the tracking camera.

[0100] Embodiment 2

[0101] See also Figure 2 In this embodiment, when using inspection drones to collect images as a basis for fault diagnosis and identification, the images collected by the image acquisition device may overlap, so as to ensure the integrity of image acquisition. The images collected by the image acquisition device can be stitched together through image stitching technology, and multiple scattered images can be stitched into a complete image with the image features of all ground equipment in the solar photovoltaic power station. On the one hand, it can prevent the duplication of fault identification due to image duplication when identifying faults, thereby affecting the efficiency and accuracy of fault identification. On the other hand, it can provide a more complete perspective and contextual information for fault identification, especially when the identification object spans multiple views, which can help the recognition algorithm to better understand the scene and object.

[0102] The specific operations are as follows:

[0103] When the backend server processes the data collected by the image acquisition device, the following steps are included:

[0104] S11: extracting and preprocessing a plurality of groups of images collected by the visible light image acquisition device, wherein the preprocessing method adopted includes image enhancement processing based on the spatial domain and image filtering processing based on the Laplace filtering method;

[0105] S12: stitching the multiple groups of images collected by the visible light image acquisition device into a total device image, wherein the total device image is one or more groups of images having images of all devices in the solar photovoltaic power station;

[0106] S13: splicing several groups of infrared images collected by the infrared image acquisition device based on the processing results of steps S11 and S12 to obtain an overall infrared image of the photovoltaic power station equipment;

[0107] S14: Process and identify the equipment visible light total image and the equipment infrared total image to identify equipment faults and defects in the solar photovoltaic power station.

[0108] Preferably, when the backend server stitches a plurality of groups of images acquired by the visible light image acquisition device, the backend server comprises the following steps:

[0109] S1201: performing distortion correction on a plurality of groups of images collected by a visible light image collection device to remove image distortion caused by a camera lens structure;

[0110] S1202: extracting sift feature points from several groups of visible light images, matching the sift feature points of two adjacent visible light images, and using the RANSAC algorithm to screen feature point pairs after matching, and eliminating erroneous feature point pairs in the feature point pairs;

[0111] S1203: Use the DLT algorithm to estimate the perspective change matrix of the remaining feature point pairs, and then use the SeamFinding algorithm to find the best stitching seam, and find the best stitching seam of the images collected by two adjacent groups of high-definition cameras; finally, use the Laplacian pyramid to decompose the original image into sub-images of different scales by differentiating the Gaussian pyramids of two adjacent layers, and perform weighted averaging on each sub-image to obtain the fusion result of each layer, and finally perform reverse reconstruction of the pyramid to obtain the final fusion effect image, while retaining the covered overlapping parts;

[0112] S1204: performing image detection on the final fusion effect image, and using an image detection algorithm based on the HSV color space model to identify illumination interference in the final fusion effect image;

[0113] S1205: Eliminate the illumination interference in the final fusion effect image, and detect the overlapping images of the eliminated part. If there is no illumination interference in the overlapping images, fill the overlapping images into the eliminated part to obtain the total visible light image of the device.

[0114] Furthermore, when several groups of infrared images collected by the infrared image acquisition device are stitched together to obtain a total infrared image of the photovoltaic power station equipment, since the infrared image may contain hot spots, and the characteristics of the hot spots in the infrared image are similar to those of the sun reflection spots in the infrared image, for the infrared image, the stitching seams of the visible light image are directly substituted into the infrared image, that is, the infrared image is stitched together through the processing results of steps S11 and S12, thereby avoiding the interference of hot spots in the infrared image.

[0115] Specifically, when the image acquisition equipment carried by the inspection drone is used to collect images of the photovoltaic power station, since the solar photovoltaic panels are relatively smooth planes, light interference such as sun reflection spots will be formed on the solar photovoltaic panels. This light interference will also affect the recognition of infrared images, which may cover up the hot spots in the illuminated area, causing missed detection of hot spots, or may be regarded as rectangular hot spots, causing false detection of hot spots. Therefore, when stitching images, images without light interference of the same object are found, and the real scene found is used as a reference to reconstruct the real scene of the light interference area, thereby achieving the effect of eliminating light interference, thereby improving the performance of image recognition and increasing the accuracy of fault detection and fault identification.

[0116] Embodiment 3

[0117] See also Figure 3 In this embodiment, when planning the inspection path of the inspection drone, the flight speed and flight altitude of the inspection drone are different under different wind speeds, and in order to ensure the clarity of the image, the flight altitude of the inspection drone needs to be adaptively adjusted according to the air visibility. Therefore, the flight parameters of the inspection drone are adjusted through the data collected by the environmental sensing module to ensure the stability of the inspection drone's operation, and the accuracy of the inspection can be guaranteed by controlling the clarity and recognizability of the image. According to the flight parameters of the inspection drone, and due to the limitations of the graphics acquisition equipment, when the inspection drone flies at different altitudes, the effective acquisition range of the graphics acquisition equipment is different. Therefore, according to the different flight parameters of the inspection drone, the inspection path of the inspection drone is re-planned to ensure the effectiveness and comprehensiveness of the inspection image acquisition. At the same time, the inspection path of the inspection drone carrying platform is re-planned, such as changing the temporary inspection stop and the driving route, which can prevent the inspection drone from having insufficient endurance and ensure the efficiency of the inspection.

[0118] Specifically, when the backend server controls the inspection drone carrying platform and the patrol drone, the following steps are included:

[0119] S21: According to the inspection area of ​​the solar photovoltaic power station, the inspection drone carrying platform is planned to determine the driving path of the inspection drone carrying platform;

[0120] S22: setting flight parameters of the inspection drone according to the various environmental parameters collected by the environmental sensing module, wherein the flight parameters of the inspection drone include flight speed, flight altitude and endurance distance;

[0121] S23: planning the inspection path of the inspection drone according to the set flight parameters of the inspection drone;

[0122] S24: According to the inspection path planning of the inspection drone, the path planning of the inspection drone carrying platform is modified.

[0123] Among them, the backend server sets the flight parameters of the inspection drone according to the various environmental parameters collected by the environmental sensor module, including the following steps:

[0124] S2201: Cleaning the data collected by various sensors in the environment monitoring module;

[0125] S2202: normalize the cleaned data, wherein the data is scaled to the interval [0,1] by using maximum and minimum normalization, wherein the principle formula is:

[0126]

[0127] in, represents the normalized data, represents the raw data samples of the sensor, Represents a manually set dataset The minimum value of Represents a manually set dataset The maximum value of

[0128] S2203: Analyze the data after data cleaning and data preprocessing to obtain the environmental interference coefficient in the solar photovoltaic power station, wherein the principle formula for analyzing the data after data cleaning and data preprocessing is:

[0129] ;

[0130] in, is the environmental interference coefficient obtained by analysis, is the value of the nth sensor after maximum and minimum normalization processing, For The nonlinear adjustment function of is a constant coefficient, where the nonlinear adjustment function Nonlinear adjustment function obtained by summarizing historical data;

[0131] S2204: Setting the flight parameters of the inspection UAV according to the environmental interference coefficient.

[0132] Furthermore, when adjusting the different flight parameters of the inspection drone, the constant coefficient The values ​​are different. For example, when the flight speed of the inspection UAV is adjusted, the constant coefficient of the light intensity parameter is 0, the constant coefficient of the wind speed parameter is 3.2, the constant coefficient of the air humidity parameter is 0.6, the constant coefficient of the air temperature parameter is 0.58, and the constant coefficient of the air visibility is 1.2; when the flight altitude of the inspection UAV is adjusted, the constant coefficient of the light intensity parameter is 0.8, the constant coefficient of the wind speed parameter is 2.1, the constant coefficient of the air humidity parameter is 0.3, the constant coefficient of the air temperature parameter is 0.78, and the constant coefficient of the air visibility is 3.5.

[0133] Furthermore, when setting the flight parameters of the inspection UAV according to the environmental interference coefficient, the corresponding flight parameters are obtained by multiplying the corresponding environmental interference coefficient with the flight parameters under ideal environmental conditions, and the flight parameters under ideal environmental conditions are determined according to the type and performance of the inspection UAV.

[0134] Example 4

[0135] See also Figure 4-Figure 8 In this embodiment, a modular cleaning device is also provided on the inspection drone carrying platform. The modular cleaning device is used to clean pollutants on the solar photovoltaic panels, and a mounting structure corresponding to the modular cleaning device is provided at the bottom of the inspection drone.

[0136] Furthermore, the modular cleaning equipment includes a mounting bracket 1, a driving motor 2, a connecting component 3, a brushing component 4 and a flushing component 5. The mounting bracket 1 is configured as a cylindrical structure, and an anti-slip layer is provided on the bottom surface of the mounting bracket 1. The driving motor 2 is provided on the top surface of the mounting bracket 1, the connecting component 3 is provided inside the mounting bracket 1, the brushing component 4 is provided inside the connecting component 3, and the flushing component 5 is provided inside the connecting component 3. A carrying component 6 is also provided on the inspection UAV carrying platform, and the carrying component 6 is used to fix and maintain the modular cleaning equipment, wherein the mounting bracket 1 is the frame and shell of the modular cleaning equipment, the driving motor 2 drives the connecting component 3 to rotate, and the connecting component 3 drives the brushing component 4 to rotate, so that the pollutants on the solar photovoltaic panel can be scrubbed, and at the same time, the flushing component 5 sprays cleaning fluid to ensure cleaning efficiency.

[0137] Furthermore, the connecting component 3 includes an output shaft 301, a connecting sleeve 302, a connecting shaft 303, a buffer spring 304 and an installation shell 305. The output shaft 301 is arranged inside the installation bracket 1, one end of the output shaft 301 is fixedly connected to the output end of the driving motor 2, the other end of the output shaft 301 is fixedly connected to the connecting sleeve 302, the buffer spring 304 is arranged inside the connecting sleeve 302, one end of the buffer spring 304 is fixedly connected to the internal top surface of the connecting sleeve 302, the other end of the buffer spring 304 is fixedly connected to the connecting shaft 303, the connecting shaft 303 is slidably connected to the connecting sleeve 302, and the installation shell 305 is arranged at the bottom end of the connecting shaft 303.

[0138] When in use, the connecting sleeve 302 is connected to the driving motor 2 through the output shaft 301, and the connecting sleeve 302 and the output shaft 301 are connected through the buffer spring 304. When the modular cleaning device is placed on the solar panel, the brushing component 4 and the flushing component 5 can be buffered to prevent damage to the solar panel or the modular cleaning device, and the brushing component 4 and the flushing component 5 can be accommodated and installed through the installation shell 305.

[0139] Furthermore, the brushing assembly 4 includes a mounting plate 401 and a fixed spring 402. The mounting plate 401 is arranged below the mounting shell 305. The fixed spring 402 is arranged in multiple groups. One end of the fixed spring 402 is fixedly connected to the bottom surface of the mounting plate 401, and the other end of the fixed spring 402 is fixedly connected to the inner top surface of the mounting shell 305. Bristles are arranged on the bottom surface of the mounting plate 401. The mounting plate 401 can be fixed by the bristles, and the mounting plate 401 can be installed by the fixed spring 402.

[0140] Furthermore, the flushing component 5 includes a storage bag 501, a connecting pipe 502, a one-way valve 503 and a nozzle 504. The storage bag 501 is arranged inside the mounting shell 305, the connecting pipe 502 is arranged at the center of the mounting plate 401, one end of the connecting pipe 502 is connected with the storage bag 501, the one-way valve 503 is arranged inside the connecting pipe 502, the nozzle 504 is arranged in multiple groups, the nozzle 504 is arranged on the mounting plate 401, and the nozzle 504 is connected with the storage bag 501.

[0141] Specifically, the storage capsule 501 can store liquids such as cleaning liquid or clean water. After the inspection drone places the modular cleaning device on the solar photovoltaic panel, the anti-skid layer on the bottom of the mounting bracket 1 increases the friction between the mounting bracket 1 and the solar photovoltaic panel, so that the mounting bracket 1 is fixed to the solar photovoltaic panel, and then the driving motor 2 drives the output shaft 301 to rotate, and the output shaft 301 drives the mounting shell 305 to rotate through the buffer spring 304, the connecting sleeve 302 and the connecting shaft 303, thereby driving the bristles on the mounting plate 401 to rotate through the fixing spring 402 to clean the solar photovoltaic panel. At the same time, after the mounting plate 401 and the bristles contact the solar photovoltaic panel, the solar photovoltaic panel squeezes the mounting plate 401, and the mounting plate 401 squeezes the storage capsule 501, so that the cleaning liquid or cleaning inside the storage capsule 501 is sprayed out through the nozzle 504, thereby cleaning the solar photovoltaic panel.

[0142] Furthermore, the carrying component 6 includes a mounting platform 601, a storage box 602, a conveying pipe 603 and a pressure pump 604. The mounting platform 601 is arranged on the inspection UAV carrying platform. A truncated cone-shaped groove is opened on the top surface of the mounting platform 601. The storage box 602 is arranged below the mounting platform 601. The conveying pipe 603 is arranged on the inner bottom surface of the truncated cone-shaped groove. The conveying pipe 603 is connected with the storage box 602, and the pressure pump 604 is arranged on the conveying pipe 603.

[0143] Specifically, the truncated cone-shaped groove on the mounting table 601 can be used to place the modular cleaning equipment into the truncated cone-shaped groove after cleaning. In combination with the cylindrical structure of the mounting bracket 1, the modular cleaning equipment can be simply returned to its original position and the connecting pipe 502 and the delivery pipe 603 can be aligned. The cleaning liquid or clean water in the storage box 602 can be pumped out by the pressure pump 604 and delivered to the storage bag 501 to replenish the liquid in the storage bag 501. This can realize the automated cleaning operation and greatly reduce labor costs.

[0144] Furthermore, the modular cleaning device also includes an auxiliary component 7, which includes a mounting plate 701, an auxiliary motor 702, a driving gear 703, a mounting shaft 704, a transmission gear 705 and a rotating arm 706. The mounting plate 701 is arranged above the mounting bracket 1, and the mounting plate 701 is fixedly connected to the mounting bracket 1. The auxiliary motor 702 is arranged on the bottom surface of the mounting plate 701, and the driving gear 703 is arranged on the output end of the auxiliary motor 702. The mounting shaft 704 is arranged in two groups, and the mounting shaft 704 is rotatably connected to the mounting plate 701. The transmission gear 705 is arranged in two groups, and the transmission gear 705 is arranged on the mounting shaft 704. The transmission gear 705 is meshed with the driving gear 703. The rotating arm 706 is arranged in two groups, and the rotating arm 706 is arranged in a semicircular structure. The rotating arm 706 is fixedly connected to the mounting shaft 704.

[0145] Furthermore, an arc-shaped groove is provided on the top surface of the mounting disk 701 , and a magnetic block is provided at the center of the top surface of the mounting disk 701 .

[0146] Furthermore, the mounting structure includes a connecting rope 801 and a hook 802 , one end of the connecting rope 801 is fixedly connected to the bottom end of the inspection drone, and the other end of the connecting rope 801 is fixedly connected to the hook 802 .

[0147] Specifically, after the inspection drone places the hook 802 on the mounting plate 701, the arc-shaped groove enables the hook 802 to move to the center of the mounting plate 701, and then the magnetic block fixes the hook 802 to prevent air flow from causing the hook 802 to move. Then the auxiliary motor 702 drives the driving gear 703 to rotate, and the driving gear 703 drives the two sets of transmission gears 705 to rotate. The transmission gear 705 drives the two sets of rotating arms 706 to rotate in the opposite direction, so that the two sets of rotating arms 706 are used to clamp the hook 802, thereby completing the mounting of the modular cleaning equipment on the inspection drone.

[0148] Example 5

[0149] Modular cleaning equipment can be used to remove pollutants detected by drone inspections, thereby preventing pollutants from affecting the power generation efficiency of solar panels and preventing the presence of pollutants from causing hot spot effects on solar panels and causing damage to the panels.

[0150] After the backend server processes and identifies the total visible light image and the total infrared image of the equipment and identifies the equipment faults and defects in the solar photovoltaic power station, the following steps are included:

[0151] S31: classifying equipment failures and defects in the solar photovoltaic power station and determining the types of equipment failures and defects;

[0152] S32: If the equipment failure and defect type is contaminants on the photovoltaic panels, the type of contaminants is identified, and different contaminant cleaning solutions are retrieved, and the retrieved contaminant cleaning solutions are sent to the intermediate processing end;

[0153] S33: The intermediate processing end controls the pressure pump 604 according to the received cleaning solution to deliver different amounts of cleaning liquid into the storage capsule 501, and controls the inspection drone to connect with the auxiliary component 7 through the mounting structure after the inspection drone completes the staged inspection target;

[0154] S34: Control the inspection drone to fly above the pollutants and descend, place the modular cleaning device above the pollutants, and clean the pollutants according to the set rotation speed and cleaning time.

[0155] Specifically, when the pollutant type is fresh bird droppings, a cleaning solution of 30% cleaning liquid and scrubbing for 20 seconds is used. When the pollutant type is dry bird droppings, a cleaning solution of 100% cleaning liquid and scrubbing for 120 seconds is used. When the pollutant type is leaves or other solid pollutants, a cleaning solution of no cleaning liquid and scrubbing for 10 seconds is used.

[0156] Specifically, when cleaning pollutants, by selecting different cleaning schemes, on the one hand, while ensuring the cleaning effect, cleaning fluid and energy consumption can be saved. On the other hand, different amounts of cleaning fluid can be delivered to the storage capsule 501 according to the type of pollutants, thereby reducing the energy consumption required for the inspection drone to carry the cleaning fluid for flight, ensuring the endurance of the inspection drone, and thus ensuring the normal progress of the inspection work.

[0157] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of those skilled in the art without departing from the spirit of the present invention.

Claims

1. A UAV system for ground inspection and fault location of solar photovoltaic power stations; characterized by: Included are: A patrol drone is used to carry image acquisition equipment to inspect the solar photovoltaic power station using the image acquisition equipment, wherein the image acquisition equipment carried includes an infrared image acquisition equipment and a visible light image acquisition equipment; The inspection drone carrying platform is used to carry the inspection drone and provide charging and transportation services for the inspection drone. The inspection drone carrying platform includes: The inspection drone lifting platform is used to provide a take-off and landing platform for the inspection drone, and the inspection drone lifting platform is provided with a mark for identification and positioning; The communication module is used to build a communication network to achieve communication between the patrol drone and the inspection drone carrying platform; The drone tracking module includes a tracking camera for tracking and collecting images of the patrol drone while it is working, and for controlling the patrol drone; The data processing end is used to process and identify the image data collected by the image acquisition device to realize the fault identification and fault location of the solar photovoltaic power station. The data processing end includes: The backend server is used to process and archive the data collected by the image acquisition equipment, and to control the inspection drone carrying platform and patrol drones; The intermediate processing end is used to process and identify the data collected by the drone tracking module and control the drone tracking module, wherein the intermediate processing end is arranged on the inspection drone carrying platform; An environmental sensing module is also provided on the inspection drone carrying platform. The environmental sensing module is used to detect whether the environment where the inspection drone carrying platform is located is suitable for the inspection drone to take off. The environmental sensing module includes a wind speed sensor, a light intensity sensor, a rain sensor and a visibility monitor. The wind speed sensor is used to detect the wind speed parameters of the environment where the inspection drone carrying platform is located. The light intensity sensor is used to detect the light intensity parameters of the environment where the inspection drone carrying platform is located. The rain sensor is used to detect the light intensity parameters of the environment where the inspection drone carrying platform is located. The visibility monitor is used to detect the visibility parameters of the environment where the inspection drone carrying platform is located. When the backend server processes the data collected by the image acquisition device, the following steps are included: S11: extracting and preprocessing a plurality of groups of images collected by the visible light image acquisition device, wherein the preprocessing method adopted includes image enhancement processing based on the spatial domain and image filtering processing based on the Laplace filtering method; S12: stitching the multiple groups of images collected by the visible light image acquisition device into a total device image, wherein the total device image is one or more groups of images having images of all devices in the solar photovoltaic power station; S13: splicing several groups of infrared images collected by the infrared image acquisition device based on the processing results of steps S11 and S12 to obtain an overall infrared image of the photovoltaic power station equipment; S14: Processing and identifying the total visible light image and the total infrared image of the equipment to identify equipment faults and defects in the solar photovoltaic power station; When the backend server stitches together a plurality of groups of images collected by the visible light image acquisition device, the following steps are included: S1201: performing distortion correction on a plurality of groups of images collected by a visible light image collection device to remove image distortion caused by a camera lens structure; S1202: extracting sift feature points from several groups of visible light images, matching the sift feature points of two adjacent visible light images, and using the RANSAC algorithm to screen feature point pairs after matching, and eliminating erroneous feature point pairs in the feature point pairs; S1203: Use the DLT algorithm to estimate the perspective change matrix of the remaining feature point pairs, and then use the SeamFinding algorithm to find the best stitching seam, and find the best stitching seam of the images collected by two adjacent groups of high-definition cameras; finally, use the Laplacian pyramid to decompose the original image into sub-images of different scales by differentiating the Gaussian pyramids of two adjacent layers, and perform weighted averaging on each sub-image to obtain the fusion result of each layer, and finally perform reverse reconstruction of the pyramid to obtain the final fusion effect image, while retaining the covered overlapping parts; S1204: performing image detection on the final fusion effect image, and using an image detection algorithm based on the HSV color space model to identify illumination interference in the final fusion effect image; S1205: Eliminate the illumination interference in the final fusion effect image, and detect the overlapping images of the eliminated part. If there is no illumination interference in the overlapping images, fill the overlapping images into the eliminated part to obtain the total visible light image of the device.

2. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 1, characterized in that: The backend server controls the inspection drone platform and patrol drones in the following steps: S21: According to the inspection area of ​​the solar photovoltaic power station, the inspection drone carrying platform is planned to determine the driving path of the inspection drone carrying platform; S22: setting flight parameters of the inspection drone according to the various environmental parameters collected by the environmental sensing module, wherein the flight parameters of the inspection drone include flight speed, flight altitude and endurance distance; S23: planning the inspection path of the inspection drone according to the set flight parameters of the inspection drone; S24: According to the inspection path planning of the inspection drone, the path planning of the inspection drone carrying platform is modified.

3. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 2 is characterized by: The backend server sets the flight parameters of the inspection drone based on the various environmental parameters collected by the environmental sensor module, including the following steps: S2201: Cleaning the data collected by various sensors in the environment monitoring module; S2202: normalize the cleaned data, wherein the maximum and minimum normalization is used to scale the data to the interval [0,1], wherein the principle formula is: in, represents the normalized data, represents the raw data samples of the sensor, Represents a manually set dataset The minimum value of Represents a manually set dataset The maximum value of S2203: Analyze the data after data cleaning and data preprocessing to obtain the environmental interference coefficient in the solar photovoltaic power station, wherein the principle formula for analyzing the data after data cleaning and data preprocessing is: ； in, is the environmental interference coefficient obtained by analysis, is the value of the nth sensor after maximum and minimum normalization processing, For The nonlinear adjustment function is is a constant coefficient, where the nonlinear adjustment function Nonlinear adjustment function obtained by summarizing historical data; S2204: Setting the flight parameters of the inspection UAV according to the environmental interference coefficient.

4. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 3 is characterized by: The drone tracking module tracks and collects images of the patrol drone while it is working, and controls the patrol drone, including the following steps: S31: After the inspection drone takes off, start the tracking camera to collect images of the inspection drone, and use a video-based image tracking algorithm to identify and track the images of the inspection drone; S32: Obtaining the location of the inspection drone according to the angle of the tracking camera, the location of the tracking camera, and the image collected by the tracking camera; S33: Control the inspection drone according to the positioning of the inspection drone, so that the inspection drone performs inspection along the set path.

5. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 4, characterized in that: A modular cleaning device is also provided on the inspection drone carrying platform, and the modular cleaning device is used to clean pollutants on the solar photovoltaic panels. A mounting structure corresponding to the modular cleaning device is provided at the bottom of the inspection drone, and the mounting structure includes a connecting rope (801) and a hook (802). One end of the connecting rope (801) is fixedly connected to the bottom end of the inspection drone, and the other end of the connecting rope (801) is fixedly connected to the hook (802).

6. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 5, characterized in that: The modular cleaning device comprises a mounting bracket (1), a driving motor (2), a connecting component (3), a scrubbing component (4) and a flushing component (5); the mounting bracket (1) is configured as a cylindrical structure, and an anti-slip layer is provided on the bottom surface of the mounting bracket (1); the driving motor (2) is provided on the top surface of the mounting bracket (1); the connecting component (3) is provided inside the mounting bracket (1); the scrubbing component (4) is provided inside the connecting component (3); the flushing component (5) is provided inside the connecting component (3); a carrying component (6) is further provided on the carrying platform of the inspection drone; the carrying component (6) is used for fixing and maintaining the modular cleaning device; the connecting component (3) comprises an output shaft (301), a connecting sleeve (302), a connecting sleeve (303), a connecting sleeve (304), a connecting sleeve (305), a connecting sleeve (306), a connecting sleeve (307), a connecting sleeve (308), a connecting sleeve (309), a connecting sleeve (310), a connecting sleeve (311), a connecting sleeve (312), a connecting sleeve (313), a connecting sleeve (314), a connecting sleeve (315), a connecting sleeve (316), a connecting sleeve (317), a connecting sleeve (318), a connecting sleeve (319), a connecting sleeve (319), a connecting sleeve (311), a connecting sleeve (319), a connecting sleeve (311), a connecting sleeve (311), a connecting sleeve (312), a connecting sleeve (313), a connecting sleeve (314), a connecting sleeve (315), a connecting sleeve (316), a connecting sleeve (317), a connecting sleeve (318), a connecting sleeve (319), a connecting sleeve (319), a connecting sleeve (311), a connecting sleeve (311), a connecting sleeve (312), a connecting sleeve (313), a connecting sleeve (314), a connecting sleeve (315), a connecting sleeve (316), a connecting sleeve (317), a connecting The invention relates to a connecting shaft (303), a buffer spring (304) and a mounting shell (305); the output shaft (301) is arranged inside the mounting bracket (1); one end of the output shaft (301) is fixedly connected to the output end of the driving motor (2); the other end of the output shaft (301) is fixedly connected to the connecting sleeve (302); the buffer spring (304) is arranged inside the connecting sleeve (302); one end of the buffer spring (304) is fixedly connected to the internal top surface of the connecting sleeve (302); the other end of the buffer spring (304) is fixedly connected to the connecting shaft (303); the connecting shaft (303) is slidably connected to the connecting sleeve (302); the mounting shell (305) is arranged at the bottom end of the connecting shaft (303); and the brushing assembly (4) comprises a mounting assembly. The mounting plate (401) and the fixing spring (402) are provided, the mounting plate (401) is arranged below the mounting shell (305), the fixing spring (402) is arranged in a plurality of groups, one end of the fixing spring (402) is fixedly connected to the bottom surface of the mounting plate (401), the other end of the fixing spring (402) is fixedly connected to the inner top surface of the mounting shell (305), the bottom surface of the mounting plate (401) is provided with bristles, the flushing assembly (5) comprises a storage capsule (501), a connecting pipe (502), a one-way valve (503) and a nozzle (504), the storage capsule (501) is arranged inside the mounting shell (305), the connecting pipe (502) is arranged at the center of the mounting plate (401), one end of the connecting pipe (502) is connected to the storage capsule, and ...) and a one-way valve (503) and a nozzle (504), the storage capsule (501) is arranged inside the mounting shell (305), the connecting pipe (502) is arranged at (501) is connected through, a one-way valve (503) is arranged inside the connecting pipe (502), a nozzle (504) is arranged in a plurality of groups, the nozzle (504) is arranged on the mounting plate (401), the nozzle (504) is connected through with the storage bag (501), the carrying component (6) includes a mounting platform (601), a storage box (602), a delivery pipeline (603) and a pressure pump (604), the mounting platform (601) is arranged on the inspection drone carrying platform, a truncated cone-shaped groove is opened on the top surface of the mounting platform (601), the storage box (602) is arranged below the mounting platform (601), the delivery pipeline (603) is arranged on the inner bottom surface of the truncated cone-shaped groove, and the delivery pipeline (603) is connected through with the storage box (602),The pressure pump (604) is arranged on the delivery pipeline (603), and the connecting pipeline (502) is arranged in alignment with the delivery pipeline (603).

7. The unmanned aerial vehicle system for ground inspection and fault location of solar photovoltaic power stations according to claim 6, characterized in that: The modular cleaning device further comprises an auxiliary component (7), wherein the auxiliary component (7) comprises a mounting plate (701), an auxiliary motor (702), a driving gear (703), a mounting shaft (704), a transmission gear (705) and a rotating arm (706), wherein the mounting plate (701) is arranged above the mounting bracket (1), the mounting plate (701) is fixedly connected to the mounting bracket (1), the auxiliary motor (702) is arranged on the bottom surface of the mounting plate (701), and the driving gear (703) is arranged on the bottom surface of the mounting plate (701). Placed on the output end of the auxiliary motor (702), the mounting shaft (704) is arranged in two groups, the mounting shaft (704) is rotatably connected to the mounting plate (701), the transmission gear (705) is arranged in two groups, the transmission gear (705) is arranged on the mounting shaft (704), the transmission gear (705) is meshed with the driving gear (703), the rotating arm (706) is arranged in two groups, the rotating arm (706) is arranged in a semicircular structure, and the rotating arm (706) is fixedly connected to the mounting shaft (704).

Images