Photovoltaic station assembly defect positioning method, device, equipment and medium

Abstract

The invention discloses a photovoltaic station assembly defect positioning method, device and equipment and a medium, and the method comprises the steps: obtaining the vertex pixel coordinates of a photovoltaic station image and the conversion relation between a photovoltaic station pixel coordinate system and a geographic coordinate system, and calculating the vertex geographic coordinates of a photovoltaic station; determining the station boundary length of the photovoltaic station based on the geographic coordinates of the vertexes, determining the number of photovoltaic station images to be split based on the station boundary length and a preset sub-station boundary length, and splitting the photovoltaic station images to obtain a plurality of photovoltaic station sub-images and the position relationship between each photovoltaic station sub-image and the photovoltaic station images; detecting component defect information in each photovoltaic station subgraph based on a pre-trained component defect detection model; and mapping the component defect information to a geographic coordinate system to obtain a component defect positioning result of the photovoltaic station. Through the technical scheme, the accuracy of a component defect judgment result and the timeliness of defect positioning can be improved.

Description

Technical Field

[0001] This application belongs to the technical field of image data processing, and particularly relates to a method, device, equipment and medium for locating component defects in a photovoltaic power station. Background Art

[0002] With the rapid development of renewable energy, as an important part of green clean energy, photovoltaic power generation has been widely used in various fields. Photovoltaic components, as the basic constituent units of a photovoltaic power station, their defect problems can directly affect the power generation efficiency of the photovoltaic power station. To ensure the operation stability of the photovoltaic power station, timely detection and location of component defects in the power station based on the image data of the photovoltaic power station have become a hot research topic in the field of photovoltaic power generation.

[0003] In the prior art, the location of component defects in a photovoltaic power station often involves pre-setting infrared imaging devices around the photovoltaic strings. By obtaining the thermal imaging data reported by the infrared imaging devices in real time, it is determined whether there are component defects in the photovoltaic components in the photovoltaic string according to whether there are abnormal temperature data in the thermal imaging data. And in the case of determining that there are component defects, the defective components are located according to the positional relationship between the infrared thermal imaging device and each photovoltaic component in the photovoltaic string.

[0004] However, the infrared thermal imaging data in the prior art is easily interfered by external environmental factors, resulting in inaccurate component defect detection results. At the same time, the prior art can only locate the approximate position of the defective components and cannot convert the specific defect information in the components into the geographic coordinate system, which is not conducive to the remote monitoring of the photovoltaic power station. Summary of the Invention

[0005] The purpose of the embodiments of this application is to provide a method, device, equipment and medium for locating component defects in a photovoltaic power station, which solves the problems of inaccurate component defect judgment results and inability to directly locate defect information in the prior art. By splitting and defect detecting the images of the photovoltaic power station, and according to the positional relationship between the split images and the original photovoltaic power station images and the conversion relationship between the image space and the geographic space, the component defect information in each sub-image of the photovoltaic power station is mapped into the geographic coordinate system to obtain the component defect location result of the photovoltaic power station, which can achieve the purpose of directly locating the component defect information of the photovoltaic power station into the geographic coordinate system, and improve the accuracy of the component defect judgment result and the timeliness of defect location.

[0006] In a first aspect, the embodiments of this application provide a method for locating component defects in a photovoltaic power station, the method includes: Obtain the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system and the geographic coordinate system of the photovoltaic power station, and calculate the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship; Determine the station boundary length of the photovoltaic power station based on the vertex geographic coordinates, determine the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length, and split the photovoltaic power station image according to the number of splits to obtain multiple photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image; Input multiple photovoltaic power station sub-images into a pre-trained component defect detection model respectively, and determine the component defect information in each photovoltaic power station sub-image based on the component defect detection model; Map the component defect information to the geographic coordinate system based on the positional relationship and the conversion relationship to obtain the component defect localization result of the photovoltaic power station.

[0007] Further, the number of splits includes the number of rows to be split and the number of columns to be split; Determining the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length includes: Perform rounding calculation on the ratio of the total width in the station boundary length to the sub-width in the preset sub-station boundary length to obtain the number of columns to be split of the photovoltaic power station image; Perform rounding calculation on the ratio of the total height in the station boundary length to the sub-height in the preset sub-station boundary length to obtain the number of rows to be split of the photovoltaic power station image.

[0008] Further, the rounding calculation is expressed by the following formula: ; where is the number of rows to be split, is the number of columns to be split, is the total height in the station boundary length, is the sub-height in the preset sub-station boundary length, is the total width in the station boundary length, is the sub-width in the preset sub-station boundary length, is the rounding algorithm.

[0009] Further, splitting the photovoltaic power station image according to the number of splits to obtain multiple photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image includes: Split the photovoltaic power station image according to the number of rows to be split and the number of columns to be split to obtain multiple photovoltaic power station sub-images; Calculate the pixel boundary length of the PV power station image based on the vertex pixel coordinates, and determine the starting vertex pixel coordinates of each PV power station sub-image in the PV power station pixel coordinate system based on the pixel boundary length, the number of rows to be split, and the number of columns to be split; Obtain the preset coincidence ratio between adjacent PV power station sub-images among multiple PV power station sub-images, and determine the ending vertex pixel coordinates of each PV power station sub-image in the PV power station pixel coordinate system based on the sub-pixel boundary length, the starting vertex pixel coordinates, and the preset coincidence ratio, so as to obtain the positional relationship between each PV power station sub-image and the PV power station image.

[0010] Further, the component defect information includes the defect type of the component defect and the initial pixel coordinates of the component defect in the PV power station sub-image pixel coordinate system; Map the component defect information to the geographic coordinate system based on the positional relationship and the conversion relationship to obtain the component defect localization result of the PV power station, including: Based on the starting vertex pixel coordinates and the initial pixel coordinates of the component defect in the PV power station sub-image pixel coordinate system, determine the final pixel coordinates of the component defect in the PV power station pixel coordinate system; Determine the defect geographic coordinates of the component defect in the geographic coordinate system based on the conversion relationship and the final pixel coordinates, and use the defect type and the defect geographic coordinates of the component defect as the component defect localization result of the PV power station.

[0011] Further, before inputting multiple PV power station sub-images into the pre-trained component defect detection model respectively, the method further includes: Obtain the standard image resolution of the pre-trained component defect detection model and the actual image resolution of each PV power station sub-image; Adjust the size of the PV power station sub-image based on the standard image resolution and the actual image resolution to obtain a standard PV power station sub-image; Correspondingly, inputting multiple PV power station sub-images into the pre-trained component defect detection model respectively includes: Input each standard PV power station sub-image into the pre-trained component defect detection model.

[0012] Further, the conversion relationship is the affine transformation matrix between the PV power station pixel coordinate system and the geographic coordinate system, where the affine transformation matrix includes the origin geographic coordinates of the origin of the PV power station pixel coordinate system in the geographic coordinate system, the pixel resolution of the PV power station image, and the rotation degree of the PV power station image; Calculate the vertex geographic coordinates of the PV power station based on the vertex pixel coordinates and the conversion relationship, including: Calculate the pixel boundary length of the PV power station image based on the vertex pixel coordinates, and calculate the vertex geographic coordinates of the PV power station based on the origin geographic coordinates, the pixel resolution, the rotation degree, and the pixel boundary length.

[0013] In a second aspect, an embodiment of the present application provides a component defect positioning device for a photovoltaic power station, and the device includes: A coordinate calculation module, configured to obtain the vertex pixel coordinates of a photovoltaic power station image and the conversion relationship between the pixel coordinate system and the geographic coordinate system of the photovoltaic power station, and calculate the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship; A position relationship determination module, configured to determine the length of the power station boundary of the photovoltaic power station based on the vertex geographic coordinates, determine the number of images of the photovoltaic power station to be split based on the length of the power station boundary and the preset length of the sub-power station boundary, and split the photovoltaic power station image according to the number of images to be split to obtain a plurality of photovoltaic power station sub-images and the position relationship between each photovoltaic power station sub-image and the photovoltaic power station image; A defect detection module, configured to input the plurality of photovoltaic power station sub-images into a pre-trained component defect detection model respectively, and determine the component defect information in each photovoltaic power station sub-image based on the component defect detection model; A defect positioning module, configured to map the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the component defect positioning result of the photovoltaic power station.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented.

[0015] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0016] In a fifth aspect, an embodiment of the present application further provides a computer program product, which includes a computer program. The computer program is stored in a computer-readable storage medium, and at least one processor of the device reads and executes the computer program, so that the device executes the method described in the first aspect.

[0017] In the embodiment of the present application, the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station and the geographic coordinate system are obtained, and the vertex geographic coordinates of the photovoltaic power station are calculated based on the vertex pixel coordinates and the conversion relationship; the station boundary length of the photovoltaic power station is determined based on the vertex geographic coordinates, the number of splits to be made for the photovoltaic power station image is determined based on the station boundary length and the preset sub-station boundary length, and the photovoltaic power station image is split according to the number of splits to be made, obtaining a plurality of photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image; the plurality of photovoltaic power station sub-images are respectively input into a pre-trained component defect detection model, and the component defect information in each photovoltaic power station sub-image is determined based on the component defect detection model; the component defect information is mapped to the geographic coordinate system based on the positional relationship and the conversion relationship, obtaining the component defect localization result of the photovoltaic power station. Through the above component defect localization method for the photovoltaic power station, the problems existing in the prior art that the component defect judgment result is not accurate enough and the defect information cannot be directly located are solved. By splitting and defect-detecting the photovoltaic power station image, and according to the positional relationship between the split image and the original photovoltaic power station image and the conversion relationship between the image space and the geographic space, the component defect information in each photovoltaic power station sub-image is mapped to the geographic coordinate system, obtaining the component defect localization result of the photovoltaic power station, which can achieve the purpose of directly locating the component defect information of the photovoltaic power station to the geographic coordinate system, improving the accuracy of the component defect judgment result and the timeliness of defect localization. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flowchart of a method for localizing component defects of a photovoltaic power station provided by an embodiment of the present application; Figure 2 is a flowchart of determining the number of splits to be made for the photovoltaic power station image provided by an embodiment of the present application; Figure 3 is a flowchart of another method for localizing component defects of a photovoltaic power station provided by an embodiment of the present application; Figure 4 is a schematic diagram of the overlapping area of the photovoltaic power station sub-images provided by the present application; Figure 5 is a structural block diagram of a device for localizing component defects of a photovoltaic power station provided by an embodiment of the present application; Figure 6 is a structural block diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] To make the objectives, technical solutions and advantages of this application clearer, the following further describes the specific embodiments of this application in conjunction with the accompanying drawings. It can be understood that the specific embodiments described herein are only used to explain this application, rather than limiting this application. Additionally, it should be noted that for ease of description, only the parts related to this application rather than all the content are shown in the drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations can be implemented in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operations are completed, but it can also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0020] The following will clearly describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application belong to the scope of protection of this application.

[0021] The terms "first", "second", etc. in the description and claims of this application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of this application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are usually of the same category, and do not limit the number of objects. For example, the first object can be one or multiple. In addition, "and / or" in the description and claims means at least one of the connected objects, and the character " / ", generally represents an "or" relationship between the associated objects before and after.

[0022] First, the usage scenario of this solution can be a scenario for locating component defects in a photovoltaic power station, especially for locating the defect positions of component defects in a photovoltaic power station to directly obtain the specific geographical locations of the component defects in the power station. By splitting and defect-detecting the photovoltaic power station images, and based on the positional relationship between the split images and the original photovoltaic power station images and the conversion relationship between the image space and the geographical space, mapping the component defect information in each sub-image of the photovoltaic power station to the geographical coordinate system, the component defect location result of the photovoltaic power station can be obtained, achieving the purpose of directly locating the component defect information of the photovoltaic power station to the geographical coordinate system, improving the accuracy of the component defect judgment result and the timeliness of defect location. Based on the above usage scenario, it can be understood that the execution subject of this solution can be an electronic device, such as intelligent terminals like mobile phones, tablets, and desktop computers.

[0023] Next, in conjunction with the accompanying drawings, a method, device, equipment, and medium for locating component defects in a photovoltaic power station provided by an embodiment of the present application will be described in detail through specific embodiments and their application scenarios.

[0024] Figure 1 is a flowchart of a method for locating component defects in a photovoltaic power station provided by an embodiment of the present application. As Figure 1 shown, it specifically includes the following steps: S101, obtain the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the photovoltaic power station pixel coordinate system and the geographical coordinate system, and calculate the vertex geographical coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship.

[0025] Among them, the photovoltaic power station image can be a remote sensing image taken of the entire photovoltaic power station. In this solution, the photovoltaic power station image is an orthophoto image of the photovoltaic power station in TIF format. The orthophoto image data includes the pixel values of the photovoltaic power station image, the geographical space coordinate reference system (CRS) where the photovoltaic power station image is located, and the mapping relationship between the pixel coordinate system and the geographical coordinate system of the photovoltaic power station image. The geographical space coordinate reference system where the photovoltaic power station image is located is used to correspond the image data of the photovoltaic power station image to the geographical space position. The vertex pixel coordinates can be the pixel coordinates of the vertex pixel points of the photovoltaic power station image in the photovoltaic power station pixel coordinate system. In this solution, the upper left vertex of the photovoltaic power station image is used as the origin of the photovoltaic power station pixel coordinate system, and its pixel coordinates are , and the vertex pixel coordinates of the upper right vertex are , and the vertex pixel coordinates of the lower left vertex are , where W is the number of width pixels of the photovoltaic power station image, and H is the number of height pixels of the photovoltaic power station image. The geographic coordinate system in this solution includes the longitude and latitude coordinate system and the projection coordinate system. After mapping the pixel coordinates in the photovoltaic power station image to the geographic space coordinate system, they are uniformly converted into the longitude and latitude coordinates in the longitude and latitude coordinate system for storage. The conversion relationship between the photovoltaic power station pixel coordinate system and the geographic coordinate system can be a conversion matrix that maps the pixel coordinates of the pixel points in the photovoltaic power station image to the corresponding longitude and latitude coordinates in the geographic space. The vertex geographic coordinates can be the longitude and latitude coordinates of the range vertices of the geographic space coverage range corresponding to the photovoltaic power station image in the geographic space.

[0026] In one embodiment, the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the photovoltaic power station pixel coordinate system and the geographic coordinate system can be obtained by reading the image data of the photovoltaic power station image. Based on the vertex pixel coordinates and the conversion relationship, the vertex pixel coordinates are mapped to the geographic space to obtain the vertex geographic coordinates of the photovoltaic power station.

[0027] In one embodiment, the conversion relationship is the affine transformation matrix between the photovoltaic power station pixel coordinate system and the geographic coordinate system. Among them, the affine transformation matrix includes the origin geographic coordinates of the origin of the photovoltaic power station pixel coordinate system in the geographic coordinate system, the pixel resolution of the photovoltaic power station image, and the rotation degree of the photovoltaic power station image; calculating the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship includes: calculating the pixel boundary length of the photovoltaic power station image based on the vertex pixel coordinates, and calculating the vertex geographic coordinates of the photovoltaic power station based on the origin geographic coordinates, pixel resolution, rotation degree, and pixel boundary length.

[0028] Among them, the affine transformation matrix is a conversion relationship formula for aligning the pixel coordinates of the photovoltaic power station image with the geographic coordinates. The origin geographic coordinates can be the longitude and latitude coordinates obtained by converting the origin pixel coordinates in the photovoltaic power station pixel coordinate system to the geographic coordinate system. The pixel resolution of the photovoltaic power station image is the actual ground distance represented by one pixel in the image. The unit of the pixel resolution of the photovoltaic power station image is meters per pixel (m / pixel). The rotation degree of the photovoltaic power station image is the angle by which the photovoltaic power station image rotates relative to the geographic coordinate axis (such as the due north direction). The unit of the rotation degree of the photovoltaic power station image is degrees (°), and the counterclockwise direction is positive. The pixel boundary length of the photovoltaic power station image can be the total number of pixels corresponding to each image side of the photovoltaic power station image. The pixel boundary length includes the number of width pixels and the number of height pixels of the image.

[0029] In one embodiment, the pixel boundary length of the photovoltaic power station image can be calculated based on the vertex pixel coordinates of two adjacent vertices on the photovoltaic power station image, and the vertex geographic coordinates of the photovoltaic power station can be calculated based on the origin geographic coordinates, pixel resolution, rotation degree, and pixel boundary length.

[0030] In one embodiment, the affine transformation matrix can be represented by the following formula: ; where, and are the geographical coordinates (longitude and latitude) of the upper left vertex of the photovoltaic power station image respectively, and are the resolutions of the pixels of the photovoltaic power station image in the x and y directions respectively (unit: meters / pixel), and represent the rotation degree of the photovoltaic power station image, usually 0.

[0031] The vertex pixel coordinates of the upper left vertex of the photovoltaic power station image can be calculated according to the origin geographical coordinates, pixel resolution, rotation degree and pixel boundary length , the vertex pixel coordinates of the upper right vertex , and the vertex pixel coordinates of the lower left vertex The corresponding longitude and latitude coordinates in the geographical coordinate system respectively, and the vertex geographical coordinates of the photovoltaic power station are obtained.

[0032] The vertex geographical coordinates of the upper left vertex of the photovoltaic power station can be calculated by the following formula: ; ; ; where LT is the vertex geographical coordinates of the upper left vertex of the photovoltaic power station, RT is the vertex geographical coordinates of the upper right vertex of the photovoltaic power station, LB is the vertex geographical coordinates of the lower left vertex of the photovoltaic power station, W is the number of pixels in the width of the photovoltaic power station image, and H is the number of pixels in the height of the photovoltaic power station image.

[0033] S102. Determine the station boundary length of the photovoltaic power station based on the vertex geographical coordinates, determine the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length, and split the photovoltaic power station image according to the number of splits to obtain a plurality of photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image.

[0034] Among them, the length of the station boundary can be the length of each side of the photovoltaic power station scope covered by the photovoltaic power station image. The preset sub-station boundary length can be the length of each side of the station scope of each sub-station after the photovoltaic power station is split into multiple sub-stations in advance. The number of splits to be made can be the number of photovoltaic power station sub-images obtained by splitting the photovoltaic power station image according to the splitting method of the sub-stations. The photovoltaic power station sub-image can be the station image corresponding to the sub-station. The positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image can be the image position of each photovoltaic power station sub-image in the photovoltaic power station image. For example: if the photovoltaic power station image is split into photovoltaic power station sub-images of M rows and N columns, the positional relationship between the i-th photovoltaic power station sub-image and the photovoltaic power station image can be expressed as .

[0035] In one embodiment, the length of the station boundary of the photovoltaic power station can be obtained by calculating the distance between two adjacent vertices according to the vertex geographical coordinates of the photovoltaic power station. If the geographical space coordinate reference system where the above photovoltaic power station image is located is a projection coordinate system, the distance between two adjacent vertices can be calculated by the following formula: ; wherein, and are respectively the vertex geographical coordinates of adjacent vertices of the photovoltaic power station in the projection coordinate system.

[0036] If the geographical space coordinate reference system where the above photovoltaic power station image is located is a longitude and latitude coordinate system, the Haversine formula can be used to calculate the distance between two adjacent longitude and latitude vertices: ; ; ; wherein, and are respectively the vertex geographical coordinates of adjacent vertices of the photovoltaic power station in the longitude and latitude coordinate system.

[0037] The ratio of the station boundary length to the preset sub-station boundary length can be calculated, and the maximum number of sub-stations included in the photovoltaic power station can be determined according to the ratio. Since the photovoltaic power station image corresponds to the geographical space scope of the entire photovoltaic power station, the maximum number of sub-stations can be used as the number of splits to be made for the photovoltaic power station image. The image area of the photovoltaic power station image is evenly split according to the number of splits to be made, and a plurality of photovoltaic power station sub-images are obtained. The positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image is determined according to the position of each photovoltaic power station sub-image in the photovoltaic power station image.

[0038] S103. Input the sub - graphs of multiple photovoltaic power stations into a pre - trained component defect detection model respectively, and determine the component defect information in each sub - graph of the photovoltaic power station based on the component defect detection model.

[0039] Among them, the component defect detection model can be a pre - trained model for detecting and locating component defects in a photovoltaic power station according to the photovoltaic power station image. For example: the YOLOv8 model. The component defect information can be the information describing the defect type of the component defect and the position of the component defect in the sub - graph of the photovoltaic power station.

[0040] In one embodiment, the sub - graphs of multiple photovoltaic power stations can be input into the pre - trained component defect detection model respectively according to the distribution order of each sub - graph in the photovoltaic power station image, and the component defect detection model is used to automatically identify the component defect information in each sub - graph of the photovoltaic power station. The component defect detection model in this solution can directly output the position coordinates of the defect box annotating the component defect in each sub - graph of the photovoltaic power station and the defect type of the defect corresponding to the defect box.

[0041] During the process of defect detection by the component defect detection model, there may be false positives (wrongly marked as defects), false negatives (missed defects), and subtle defect features. In this solution, the CBAM channel attention mechanism is introduced into the YOLOv8 model. CBAM consists of two main parts, namely the channel attention module and the spatial attention module. After the two are combined, the false detection and missed detection situations can be effectively reduced. At the same time, in this solution, a small - object detection head is added to the Head layer of the YOLOv8 network structure and combined with the CBAM mechanism to further improve the model's detection ability for subtle defects of photovoltaic components.

[0042] Before training the component defect detection model, a large number of photovoltaic component images can be collected in advance, and the LabelImg tool is used to annotate the photovoltaic component images, marking defect types such as dot - like hot spots, strip - like hot spots, vegetation occlusion, snow cover, and photovoltaic panel missing, etc., to make a data set. The OpenCV is used to clean the data set, removing incorrect or irrelevant data, and enhancing the data set through operations such as rotation, cropping, flipping, and translation to improve the accuracy and generalization of the model. After the model structure of the component defect detection model is constructed, the above data set is divided into a training set, a test set, and a validation set according to 8:1:1 respectively. Among them, the test set and the validation set are randomly sampled from the original data set to ensure data diversity. Finally, the YOLOv8 model is imported for defect detection training, and after adjusting some network hyperparameters, the training for 300 epochs is started to obtain the final component defect detection model.

[0043] In one embodiment, before separately inputting multiple sub - graphs of photovoltaic power stations into a pre - trained component defect detection model, the method further includes: obtaining the standard image resolution of the pre - trained component defect detection model and the actual image resolution of each sub - graph of the photovoltaic power station; adjusting the size of the sub - graph of the photovoltaic power station based on the standard image resolution and the actual image resolution to obtain a standard sub - graph of the photovoltaic power station; correspondingly, separately inputting multiple sub - graphs of the photovoltaic power station into the pre - trained component defect detection model includes: inputting each standard sub - graph of the photovoltaic power station into the pre - trained component defect detection model.

[0044] Among them, the standard image resolution can be a parameter used to describe the minimum clarity of the input image required by the component defect detection model. The actual image resolution can be a parameter used to describe the actual clarity of each sub - graph of the photovoltaic power station. The standard sub - graph of the photovoltaic power station can be a sub - graph of the photovoltaic power station whose clarity meets the standard image resolution.

[0045] In one embodiment, the standard image resolution of the pre - trained component defect detection model and the actual image resolution of each sub - graph of the photovoltaic power station can be obtained, the scaling ratio for resizing the sub - graph of the photovoltaic power station can be determined according to the ratio of the standard image resolution to the actual image resolution, the size of the sub - graph of the photovoltaic power station can be adjusted according to the scaling ratio to obtain a standard sub - graph of the photovoltaic power station that meets the model requirements, and each standard sub - graph of the photovoltaic power station can be input into the pre - trained component defect detection model to accurately detect and locate the component defects in the sub - graph of the photovoltaic power station by using the component defect detection model.

[0046] S104, map the component defect information to the geographic coordinate system based on the position relationship and the transformation relationship to obtain the component defect location result of the photovoltaic power station.

[0047] Among them, the component defect location result can be the position information of the component defect in the photovoltaic power station in the geographical space.

[0048] In one embodiment, the component defect information in each sub - graph of the photovoltaic power station can be mapped to the entire photovoltaic power station image according to the position relationship, and then the component defect information in the photovoltaic power station image can be mapped to the geographic coordinate system according to the transformation relationship to obtain the component defect location result of the photovoltaic power station.

[0049] The technical solution provided by the embodiments of the present application obtains the vertex pixel coordinates of a photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station and the geographic coordinate system, calculates the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship; determines the station boundary length of the photovoltaic power station based on the vertex geographic coordinates, determines the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length, and splits the photovoltaic power station image according to the number of splits to obtain a plurality of photovoltaic power station sub-images and the position relationship between each photovoltaic power station sub-image and the photovoltaic power station image; inputs the plurality of photovoltaic power station sub-images into a pre-trained component defect detection model respectively, and determines the component defect information in each photovoltaic power station sub-image based on the component defect detection model; maps the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the component defect positioning result of the photovoltaic power station. Through the above component defect positioning method of the photovoltaic power station, the problems existing in the prior art that the component defect judgment result is not accurate enough and the defect information cannot be directly located are solved. By splitting and defect detecting the photovoltaic power station image, and according to the position relationship between the split image and the original photovoltaic power station image and the conversion relationship between the image space and the geographic space, the component defect information in each photovoltaic power station sub-image is mapped to the geographic coordinate system to obtain the component defect positioning result of the photovoltaic power station, which can achieve the purpose of directly positioning the component defect information of the photovoltaic power station to the geographic coordinate system, and improve the accuracy of the component defect judgment result and the timeliness of defect positioning.

[0050] Figure 2 It is a flowchart for determining the number of splits of a photovoltaic power station image provided by the embodiments of the present application. The number of splits includes the number of rows to be split and the number of columns to be split. As Figure 2 shown, it specifically includes the following steps: S201, perform integer rounding calculation on the ratio of the total width in the station boundary length to the sub-width in the preset sub-station boundary length to obtain the number of columns to be split of the photovoltaic power station image.

[0051] Among them, the station boundary includes the width boundary and the height boundary of the photovoltaic power station. The total width in the station boundary length may be the boundary length value of the width boundary of the photovoltaic power station. The sub-width in the preset sub-station boundary length may be the boundary length value of the width boundary of each pre-set sub-station. The integer rounding calculation may include upward rounding calculation and rounding calculation. The number of columns to be split may be the total number of columns required to split the photovoltaic power station image.

[0052] In one embodiment, the ratio of the total width in the station boundary length to the sub-width in the preset sub-station boundary length may be calculated, and the ratio may be rounded to obtain the number of columns to be split of the photovoltaic power station image.

[0053] S202. Perform rounding calculation on the ratio of the total height in the boundary length of the station yard to the sub-height in the preset sub-station yard boundary length to obtain the number of rows to be split for the photovoltaic power station image.

[0054] Among them, the total height in the boundary length of the station yard can be the boundary length value of the high boundary of the photovoltaic power station. The sub-height in the preset sub-station yard boundary length can be the boundary length value of the high boundary of each preset sub-station yard. The number of rows to be split can be the total number of rows required to split the photovoltaic power station image.

[0055] In one embodiment, the ratio of the total height in the boundary length of the station yard to the sub-height in the preset sub-station yard boundary length can be calculated, and the ratio can be rounded to obtain the number of rows to be split for the photovoltaic power station image.

[0056] In one embodiment, the rounding calculation of the ratio is represented by the following formula: ; Among them, is the number of rows to be split, is the number of columns to be split, is the total height in the boundary length of the station yard, is the sub-height in the preset sub-station yard boundary length, is the total width in the boundary length of the station yard, is the sub-width in the preset sub-station yard boundary length, is the rounding algorithm.

[0057] In one embodiment, the rounding calculation of the ratio can be the rounding algorithm, which is represented by the following formula: ; In the case where the ratio of the total height in the boundary length of the station yard to the sub-height in the preset sub-station yard boundary length is less than 1, the number of split rows is taken as 1. In the case where the ratio of the total width in the boundary length of the station yard to the sub-width in the preset sub-station yard boundary length is less than 1, the number of split columns is taken as 1.

[0058] In this solution, the rounding algorithm of the ratio ensures that the number of rows and columns for splitting the photovoltaic power station image is at least 1, avoiding the problem that the preset boundary value of the sub-graph of the photovoltaic power station is too large to calculate the splitting quantity of the photovoltaic power station.

[0059] The technical solution provided by the embodiments of the present application calculates the integer ratio of the total width in the field boundary length to the sub-width in the preset sub-field boundary length to obtain the number of columns to be split for the photovoltaic power station image, and calculates the integer ratio of the total height in the field boundary length to the sub-height in the preset sub-field boundary length to obtain the number of rows to be split for the photovoltaic power station image, which can improve the rationality of splitting the photovoltaic power station image and is beneficial to improving the accuracy of subsequent component defect detection for the sub-images of the photovoltaic power station image.

[0060] Figure 3 It is a flowchart of another method for locating component defects in a photovoltaic power station provided by an embodiment of the present application. As Figure 3 shown, it specifically includes the following steps: S301, obtain the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station and the geographic coordinate system, and calculate the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship.

[0061] S302, determine the field boundary length of the photovoltaic power station based on the vertex geographic coordinates, and calculate the integer ratio of the total width in the field boundary length to the sub-width in the preset sub-field boundary length to obtain the number of columns to be split for the photovoltaic power station image.

[0062] S303, calculate the integer ratio of the total height in the field boundary length to the sub-height in the preset sub-field boundary length to obtain the number of rows to be split for the photovoltaic power station image.

[0063] S304, split the photovoltaic power station image according to the number of rows to be split and the number of columns to be split to obtain multiple sub-images of the photovoltaic power station.

[0064] Among them, the sub-image of the photovoltaic power station can be an image obtained by splitting the photovoltaic power station image.

[0065] In one embodiment, the photovoltaic power station image can be preliminarily split according to the number of rows to be split and the number of columns to be split to obtain multiple sub-images of the photovoltaic power station.

[0066] S305, calculate the pixel boundary length of the photovoltaic power station image based on the vertex pixel coordinates, and determine the starting vertex pixel coordinates of each sub-image of the photovoltaic power station in the pixel coordinate system of the photovoltaic power station based on the pixel boundary length, the number of rows to be split, and the number of columns to be split.

[0067] Among them, the pixel boundary length can be the total pixel length of each boundary of the photovoltaic power station image. The pixel boundary length includes the total pixel length of the wide boundary of the photovoltaic power station image and the total pixel length of the high boundary. The starting vertex pixel coordinates include the vertex pixel coordinates of the upper left corner vertex of the upper boundary of the sub-image of the photovoltaic power station.

[0068] In one embodiment, the pixel distance between two adjacent vertices can be calculated based on the vertex pixel coordinates of the two adjacent vertices to obtain the pixel boundary length of the photovoltaic power station image. According to the total length of the wide boundary pixels and the number of columns to be split, the starting vertex pixel coordinates of the first row of photovoltaic power station sub-images in the photovoltaic power station pixel coordinate system can be calculated. According to the total length of the high boundary pixels and the number of rows to be split, the starting vertex pixel coordinates of each row of photovoltaic power station sub-images below the first row in the photovoltaic power station pixel coordinate system can be calculated.

[0069] S306. Obtain the preset overlapping ratio between adjacent photovoltaic power station sub-images among multiple photovoltaic power station sub-images, and determine the termination vertex pixel coordinates of each photovoltaic power station sub-image in the photovoltaic power station pixel coordinate system based on the sub-pixel boundary length, the starting vertex pixel coordinates, and the preset overlapping ratio, so as to obtain the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image.

[0070] Among them, the sub-pixel boundary length can be the pixel boundary length of a preset sub-power station in the pixel coordinate system. The sub-pixel boundary length can be determined by calculating the total number of pixels corresponding to the preset sub-power station boundary length, and can also be determined by calculating the ratio of the pixel boundary length of the photovoltaic power station image to the number of rows to be split and the ratio of the pixel boundary length to the number of columns to be split. The preset overlapping ratio can be a parameter for the overlapping degree of the images between adjacent photovoltaic power station sub-images set in advance. The termination vertex pixel coordinates can be the vertex pixel coordinates obtained by adding the preset overlapping ratio to the original termination vertex coordinates after each photovoltaic power station sub-image is split according to the number of rows to be split and the number of columns to be split.

[0071] Figure 4 It is a schematic diagram of the overlapping area of the photovoltaic power station sub-image provided by this application.

[0072] As Figure 4 shown, the A area in the figure is the photovoltaic power station sub-image. The part enclosed by the dotted line is the overlapping area between the A area in the figure and the adjacent sub-image on the right. The upper left corner vertex is the starting vertex pixel coordinate of the A area, and the lower right corner vertex is the termination vertex pixel coordinate of the A area. To avoid the problem that the same components are split after the photovoltaic power station image is split into photovoltaic power station sub-images, resulting in the subsequent component defect detection model being unable to accurately identify the component defects in the photovoltaic power station sub-images, there is a certain overlapping area between adjacent photovoltaic power station sub-images in this solution. The size of the overlapping area is the preset overlapping ratio, and the preset overlapping ratio in this solution is 10% of the size of each sub-image.

[0073] In one embodiment, a preset overlap ratio between adjacent photovoltaic power station sub - graphs among multiple photovoltaic power station sub - graphs can be obtained. The original start and end vertex coordinates of each photovoltaic power station sub - graph can be determined according to the sub - pixel boundary length and the starting vertex pixel coordinates. The overlapping pixel length that each photovoltaic power station sub - graph needs to overlap with the adjacent photovoltaic power station sub - graph can be determined according to the sub - pixel boundary length and the preset overlap ratio. The start and end vertex pixel coordinates in the photovoltaic power station pixel coordinate system of each photovoltaic power station sub - graph can be obtained by adding the original start and end vertex coordinates and the overlapping pixel length. Since the photovoltaic power station pixel coordinate system in this solution is constructed based on the photovoltaic power station image and the coordinate origin coincides with the upper - left vertex of the photovoltaic power station image, therefore, the starting vertex pixel coordinates and the ending vertex pixel coordinates of each photovoltaic power station sub - graph can be used as the positional relationship between each photovoltaic power station sub - graph and the photovoltaic power station image.

[0074] In one embodiment, the sub - pixel boundary length can be calculated by the following formula: ; ; where, is the width pixel boundary length in the sub - pixel boundary length, is the height pixel boundary length in the sub - pixel boundary length, is the number of columns to be split, is the number of rows to be split, and determine how many sub - graphs the photovoltaic power station image needs to be split into.

[0075] The starting vertex pixel coordinates of the photovoltaic power station sub - graph in the i - th row and j - th column can be calculated according to the following formula: ; ; ; ; where, i is the row number of the photovoltaic power station sub - graph on the photovoltaic power station image, j is the column number of the photovoltaic power station sub - graph on the photovoltaic power station image, is the pixel coordinate value on the x - axis of the starting vertex pixel coordinates of the upper - left vertex of the photovoltaic power station sub - graph in the i - th row and j - th column, is the pixel coordinate value on the y - axis of the starting vertex pixel coordinates of the upper - left vertex of the photovoltaic power station sub - graph in the i - th row and j - th column, is the i - th row, the j - th column.

[0076] The ending vertex pixel coordinates of the photovoltaic power station sub - graph in the i - th row and j - th column can be calculated according to the following formula: ; ; Wherein, is the pixel coordinate value on the x-axis of the termination vertex pixel coordinate of the lower right vertex of the photovoltaic power station sub-graph in the i-th row and j-th column, is the pixel coordinate value on the y-axis of the termination vertex pixel coordinate of the lower right vertex of the photovoltaic power station sub-graph in the i-th row and j-th column.

[0077] S307. Input the multiple photovoltaic power station sub-graphs into a pre-trained component defect detection model respectively, and determine the component defect information in each photovoltaic power station sub-graph based on the component defect detection model.

[0078] S308. Map the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the component defect localization result of the photovoltaic power station.

[0079] In one embodiment, the component defect information includes the defect type of the component defect and the initial pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station sub-graph; mapping the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the component defect localization result of the photovoltaic power station, including: determining the final pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station based on the starting vertex pixel coordinates and the initial pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station sub-graph; determining the defect geographic coordinates of the component defect in the geographic coordinate system based on the conversion relationship and the final pixel coordinates, and taking the defect type and the defect geographic coordinates of the component defect as the component defect localization result of the photovoltaic power station.

[0080] Wherein, the defect type may include dot hot spots, strip hot spots, vegetation occlusion, snow cover, and photovoltaic panel missing, etc.

[0081] In one embodiment, the sum of the starting vertex pixel coordinates of the photovoltaic power station sub-graph and the initial pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station sub-graph can be calculated to obtain the final pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station, and the defect geographic coordinates of the component defect in the geographic coordinate system are determined according to the conversion relationship and the final pixel coordinates, and the defect type and the defect geographic coordinates of the component defect are taken as the component defect localization result of the photovoltaic power station.

[0082] In one embodiment, the row and column indexes of each photovoltaic power station sub-graph in the photovoltaic power station image can be recorded , obtain the defect information for component defect detection of the photovoltaic power station sub - graph output by the YOLOv8 model, and get the component defect categories and defect box coordinates in the photovoltaic power station sub - graph. Since the defect box coordinates are relative to the scaled - down sub - graph, they need to be converted back to the pixel coordinates in the original sub - graph and then mapped to the photovoltaic power station pixel coordinate system of the entire photovoltaic power station image. The coordinates for mapping the defect box coordinates to the photovoltaic power station image coordinate system can be calculated by the following formula: ; ; Among them, is the coordinate value on the x - axis for mapping the defect box coordinates to the photovoltaic power station image coordinate system, is the coordinate value on the y - axis for mapping the defect box coordinates to the photovoltaic power station image coordinate system, is the coordinate value on the x - axis of the defect box coordinates in the photovoltaic power station sub - image pixel coordinate system, is the coordinate value on the y - axis of the defect box coordinates in the photovoltaic power station sub - image pixel coordinate system.

[0083] The pixel coordinates in the photovoltaic power station image coordinate system can be mapped to the geographic coordinate system by the following formula: ; Among them, T is the conversion relationship between the above - mentioned photovoltaic power station pixel coordinate system and the geographic coordinate system, is the pixel coordinate of the sub - graph in the i - th row and j - th column in the photovoltaic power station image in the photovoltaic power station pixel coordinate system, is the above - mentioned geographic coordinate corresponding to the pixel position.

[0084] In order to unify the coordinate storage format of geographic coordinates and facilitate subsequent component defect analysis, the geographic coordinates that are not in the EPSG:4326 (WGS84) longitude - latitude format in the geographic coordinates need to be converted to the longitude - latitude format by the following formula: ; Among them, is the converted WGS84 geographic coordinate (longitude - latitude), is the original coordinate system code (such as UTM32650), represents the WGS84 coordinate system, represents the transformation function from the original coordinate system to the WGS84 coordinate system.

[0085] In this solution, based on the starting vertex pixel coordinates and the initial pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station sub-image, the final pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station are determined. Based on the conversion relationship and the final pixel coordinates, the defect geographical coordinates of the component defect in the geographical coordinate system are determined. Taking the defect type and defect geographical coordinates of the component defect as the component defect positioning result of the photovoltaic power station can achieve the purpose of accurately mapping the component defect information to the geographical coordinate system, improving the timeliness of component defect positioning.

[0086] In the technical solution provided by the embodiment of the present application, the pixel boundary length of the photovoltaic power station image is calculated based on the vertex pixel coordinates. Based on the pixel boundary length, the number of rows to be split, and the number of columns to be split, the starting vertex pixel coordinates of each photovoltaic power station sub-image in the pixel coordinate system of the photovoltaic power station are determined. Based on the sub-pixel boundary length, the starting vertex pixel coordinates, and the preset coincidence ratio, the ending vertex pixel coordinates of each photovoltaic power station sub-image in the pixel coordinate system of the photovoltaic power station are determined, obtaining the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image, which can avoid the problem of splitting the same component in the photovoltaic power station image by the photovoltaic power station sub-image and improve the accuracy of the component defect detection result.

[0087] Figure 5 It is the structural block diagram of a component defect positioning device for a photovoltaic power station provided by an embodiment of the present application. As Figure 5 shown, it specifically includes the following: The coordinate calculation module 501 is used to obtain the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station and the geographical coordinate system, and calculate the vertex geographical coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship; The position relationship determination module 502 is used to determine the station boundary length of the photovoltaic power station based on the vertex geographical coordinates, determine the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length, split the photovoltaic power station image according to the number of splits, obtain a plurality of photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image; The defect detection module 503 is used to input the plurality of photovoltaic power station sub-images into a pre-trained component defect detection model respectively, and determine the component defect information in each photovoltaic power station sub-image based on the component defect detection model; The defect positioning module 504 is used to map the component defect information to the geographical coordinate system based on the position relationship and the conversion relationship, and obtain the component defect positioning result of the photovoltaic power station.

[0088] Furthermore, the number of splits includes the number of rows to be split and the number of columns to be split; The position relationship determination module 502 is specifically used for: Perform rounding calculation on the ratio of the total width in the station boundary length to the sub-width in the preset sub-station boundary length to obtain the number of columns to be split for the photovoltaic power station image; Perform rounding calculation on the ratio of the total height in the station boundary length to the sub-height in the preset sub-station boundary length to obtain the number of rows to be split for the photovoltaic power station image.

[0089] Furthermore, the rounding calculation is represented by the following formula: ; where, is the number of rows to be split, is the number of columns to be split, is the total height in the station boundary length, is the sub-height in the preset sub-station boundary length, is the total width in the station boundary length, is the sub-width in the preset sub-station boundary length, is the rounding algorithm.

[0090] Furthermore, the position relationship determination module 502 is specifically used for: Split the photovoltaic power station image according to the number of rows to be split and the number of columns to be split to obtain multiple photovoltaic power station sub-images; Calculate the pixel boundary length of the photovoltaic power station image based on the vertex pixel coordinates, and determine the starting vertex pixel coordinates of each photovoltaic power station sub-image in the photovoltaic power station pixel coordinate system based on the pixel boundary length, the number of rows to be split, and the number of columns to be split; Obtain the preset overlap ratio between adjacent photovoltaic power station sub-images among the multiple photovoltaic power station sub-images, and determine the ending vertex pixel coordinates of each photovoltaic power station sub-image in the photovoltaic power station pixel coordinate system based on the sub-pixel boundary length, the starting vertex pixel coordinates, and the preset overlap ratio, so as to obtain the position relationship between each photovoltaic power station sub-image and the photovoltaic power station image.

[0091] Furthermore, the component defect information includes the defect type of the component defect and the initial pixel coordinates of the component defect in the photovoltaic power station sub-image pixel coordinate system; The defect location module 504 is specifically used for: Based on the starting vertex pixel coordinates and the initial pixel coordinates of the component defect in the photovoltaic power station sub-image pixel coordinate system, determine the final pixel coordinates of the component defect in the photovoltaic power station pixel coordinate system; Based on the conversion relationship and the final pixel coordinates, determine the defect geographical coordinates of the component defect in the geographical coordinate system, and use the defect type and the defect geographical coordinates of the component defect as the component defect location result of the photovoltaic power station.

[0092] Furthermore, the device further includes: A resolution acquisition module is configured to acquire the standard image resolution of a pre-trained component defect detection model and the actual image resolution of each sub-graph of a photovoltaic power station yard. A size adjustment module is configured to adjust the size of the sub-graph of the photovoltaic power station yard based on the standard image resolution and the actual image resolution to obtain a standard sub-graph of the photovoltaic power station yard. Correspondingly, the defect detection module 503 is specifically configured to: Input each standard sub-graph of the photovoltaic power station yard into the pre-trained component defect detection model.

[0093] Furthermore, the conversion relationship is the affine transformation matrix between the pixel coordinate system of the photovoltaic power station yard and the geographic coordinate system, where the affine transformation matrix includes the origin geographic coordinates of the origin of the pixel coordinate system of the photovoltaic power station yard in the geographic coordinate system, the pixel resolution of the photovoltaic power station yard image, and the rotation degree of the photovoltaic power station yard image. The coordinate calculation module 501 is specifically configured to: Calculate the pixel boundary length of the photovoltaic power station yard image based on the vertex pixel coordinates, and calculate the vertex geographic coordinates of the photovoltaic power station yard based on the origin geographic coordinates, pixel resolution, rotation degree, and pixel boundary length.

[0094] In the technical solution provided by the embodiments of the present application, the coordinate calculation module is configured to acquire the vertex pixel coordinates of the photovoltaic power station yard image and the conversion relationship between the pixel coordinate system of the photovoltaic power station yard and the geographic coordinate system, and calculate the vertex geographic coordinates of the photovoltaic power station yard based on the vertex pixel coordinates and the conversion relationship; the position relationship determination module is configured to determine the yard boundary length of the photovoltaic power station yard based on the vertex geographic coordinates, determine the number of sub-yard images to be split of the photovoltaic power station yard image based on the yard boundary length and the preset sub-yard boundary length, and split the photovoltaic power station yard image according to the number of sub-yard images to be split to obtain a plurality of sub-graphs of the photovoltaic power station yard and the position relationship between each sub-graph of the photovoltaic power station yard and the photovoltaic power station yard image; the defect detection module is configured to input the plurality of sub-graphs of the photovoltaic power station yard into the pre-trained component defect detection model respectively, and determine the component defect information in each sub-graph of the photovoltaic power station yard based on the component defect detection model; the defect location module is configured to map the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the component defect location result of the photovoltaic power station yard. Through the above-mentioned component defect location device for a photovoltaic power station yard, the problems existing in the prior art, such as inaccurate component defect judgment results and inability to directly locate defect information, are solved. By splitting and defect-detecting the photovoltaic power station yard image, and according to the position relationship between the split image and the original photovoltaic power station yard image and the conversion relationship between the image space and the geographic space, the component defect information in each sub-graph of the photovoltaic power station yard is mapped to the geographic coordinate system to obtain the component defect location result of the photovoltaic power station yard, which can achieve the purpose of directly locating the component defect information of the photovoltaic power station yard to the geographic coordinate system, and improve the accuracy of the component defect judgment result and the timeliness of defect location.

[0095] A component defect location device in an embodiment of the present application can be configured in a device, or can be configured in components, integrated circuits, or chips in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. Exemplarily, the mobile electronic device can be a mobile phone, a tablet computer, a laptop computer, a palmtop computer, an in-vehicle electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc., and the non-mobile electronic device can be a server, a Network Attached Storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc. The embodiments of the present application do not make specific limitations.

[0096] A component defect location device in an embodiment of the present application can be an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems. The embodiments of the present application do not make specific limitations.

[0097] A component defect location device provided in an embodiment of the present application can implement each process implemented in the above method embodiments. To avoid repetition, it will not be elaborated here.

[0098] As Figure 6 shown, an embodiment of the present application further provides an electronic device 600, including a processor 601, a memory 602, and a program or instruction stored on the memory 602 and executable on the processor 601. When the program or instruction is executed by the processor 601, it implements each process of the above method embodiment of a component defect location method for a photovoltaic power station, and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0099] It should be noted that the electronic device in the embodiments of the present application includes the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0100] An embodiment of the present application further provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, it implements each process of the above method embodiment of a component defect location method for a photovoltaic power station, and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0101] Among them, the processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs, etc.

[0102] Another embodiment of the present application provides a program product, which includes program code. When the program product runs on a computer device, the program code is used to cause the computer device to execute the steps in the methods according to various exemplary embodiments of the present application described above. For example, the computer device can execute a method for locating component defects in a photovoltaic power station described in an embodiment of the present application. The program product can be implemented by any combination of one or more readable media.

[0103] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article, or device including that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the methods described can be performed in an order different from that described, and various steps can also be added, omitted, or combined. Additionally, features described with reference to certain examples can be combined in other examples.

[0104] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc) and includes several instructions for causing a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present application.

[0105] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms without departing from the purpose of the present application and the scope protected by the claims, and all of them fall within the protection scope of the present application.

[0106] The above is only the preferred embodiment of the present application and the technical principles applied. The present application is not limited to the specific embodiments described herein. Various obvious changes, re-adjustments and substitutions that can be made by those skilled in the art will not depart from the protection scope of the present application. Therefore, although the present application has been described in more detail through the above embodiments, the present application is not limited to the above embodiments. Without departing from the concept of the present application, more other equivalent embodiments can be included, and the scope of the present application is determined by the scope of the claims.

Claims

1. A method for locating component defects in a photovoltaic power station, characterized in that, The method includes: Obtaining the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station and the geographic coordinate system, and calculating the vertex geographic coordinates of the photovoltaic power station based on the vertex pixel coordinates and the conversion relationship; Determining the station boundary length of the photovoltaic power station based on the vertex geographic coordinates, determining the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length, and splitting the photovoltaic power station image according to the number of splits to obtain a plurality of photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image; Inputting the plurality of photovoltaic power station sub-images into a pre-trained component defect detection model respectively, and determining the component defect information in each photovoltaic power station sub-image based on the component defect detection model; Mapping the component defect information to the geographic coordinate system based on the positional relationship and the conversion relationship to obtain the component defect localization result of the photovoltaic power station.

2. The method for locating component defects in a photovoltaic power station according to claim 1, characterized in that, The number of splits includes the number of rows to be split and the number of columns to be split; The determining the number of splits of the photovoltaic power station image based on the station boundary length and the preset sub-station boundary length includes: Performing an integer ratio calculation on the total width in the station boundary length and the sub-width in the preset sub-station boundary length to obtain the number of columns to be split of the photovoltaic power station image; Performing an integer ratio calculation on the total height in the station boundary length and the sub-height in the preset sub-station boundary length to obtain the number of rows to be split of the photovoltaic power station image.

3. The method for locating component defects in a photovoltaic power station according to claim 2, characterized in that, The integer ratio calculation is represented by the following formula: ； Among them, is the number of rows to be split, is the number of columns to be split, is the total height in the length of the station boundary, is the sub-height in the length of the preset sub-station boundary, is the total width in the length of the station boundary, is the sub-width in the length of the preset sub-station boundary, is the rounding algorithm.

4. The method for locating component defects in a photovoltaic power station according to claim 2, characterized in that, The splitting the photovoltaic power station image according to the number of splits to obtain a plurality of photovoltaic power station sub-images and the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image includes: Splitting the photovoltaic power station image according to the number of rows to be split and the number of columns to be split to obtain a plurality of photovoltaic power station sub-images; Calculating the pixel boundary length of the photovoltaic power station image based on the vertex pixel coordinates, and determining the starting vertex pixel coordinates of each photovoltaic power station sub-image in the pixel coordinate system of the photovoltaic power station based on the pixel boundary length, the number of rows to be split, and the number of columns to be split; Obtaining the preset overlapping ratio between adjacent photovoltaic power station sub-images among the plurality of photovoltaic power station sub-images, and determining the ending vertex pixel coordinates of each photovoltaic power station sub-image in the pixel coordinate system of the photovoltaic power station based on the sub-pixel boundary length, the starting vertex pixel coordinates, and the preset overlapping ratio to obtain the positional relationship between each photovoltaic power station sub-image and the photovoltaic power station image.

5. The method for locating component defects in a photovoltaic power station according to claim 4, characterized in that, The component defect information includes the defect type of the component defect and the initial pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station sub-image; The mapping the component defect information to the geographic coordinate system based on the positional relationship and the conversion relationship to obtain the component defect localization result of the photovoltaic power station includes: Determine the final pixel coordinates of the component defect in the pixel coordinate system of the photovoltaic power station yard based on the starting vertex pixel coordinates and the initial pixel coordinates of the component defect in the sub-pixel coordinate system of the photovoltaic power station yard; Determine the defective geographic coordinates of the component defect in the geographic coordinate system based on the conversion relationship and the final pixel coordinates, and use the defect type and the defective geographic coordinates of the component defect as the positioning result of the component defect in the photovoltaic power station yard.

6. The method for locating component defects in a photovoltaic power station according to claim 1, characterized in that, Before separately inputting the multiple photovoltaic power station sub-images into a pre-trained component defect detection model, the method further includes: Obtain the standard image resolution of the pre-trained component defect detection model and the actual image resolution of each photovoltaic power station sub-image; Adjust the size of the photovoltaic power station sub-image based on the standard image resolution and the actual image resolution to obtain a standard photovoltaic power station sub-image; Correspondingly, the step of separately inputting the multiple photovoltaic power station sub-images into a pre-trained component defect detection model includes: Input each standard photovoltaic power station sub-image into a pre-trained component defect detection model.

7. The method for locating component defects in a photovoltaic power station according to claim 1, characterized in that, The conversion relationship is the affine transformation matrix between the pixel coordinate system of the photovoltaic power station yard and the geographic coordinate system, where the affine transformation matrix includes the origin geographic coordinates of the origin of the pixel coordinate system of the photovoltaic power station yard in the geographic coordinate system, the pixel resolution of the photovoltaic power station image, and the rotation degree of the photovoltaic power station image; The step of calculating the vertex geographic coordinates of the photovoltaic power station yard based on the vertex pixel coordinates and the conversion relationship includes: Calculate the pixel boundary length of the photovoltaic power station image based on the vertex pixel coordinates, and calculate the vertex geographic coordinates of the photovoltaic power station yard based on the origin geographic coordinates, the pixel resolution, the rotation degree, and the pixel boundary length.

8. A device for locating component defects in a photovoltaic power station, characterized in that, The device includes: A coordinate calculation module, configured to obtain the vertex pixel coordinates of the photovoltaic power station image and the conversion relationship between the pixel coordinate system of the photovoltaic power station yard and the geographic coordinate system, and calculate the vertex geographic coordinates of the photovoltaic power station yard based on the vertex pixel coordinates and the conversion relationship; A position relationship determination module, configured to determine the yard boundary length of the photovoltaic power station yard based on the vertex geographic coordinates, determine the number of sub-yards to be split of the photovoltaic power station image based on the yard boundary length and a preset sub-yard boundary length, and split the photovoltaic power station image according to the number of sub-yards to be split to obtain a plurality of photovoltaic power station sub-images and the position relationship between each photovoltaic power station sub-image and the photovoltaic power station image; A defect detection module, configured to separately input the multiple photovoltaic power station sub-images into a pre-trained component defect detection model, and determine the component defect information in each photovoltaic power station sub-image based on the component defect detection model; A defect positioning module, configured to map the component defect information to the geographic coordinate system based on the position relationship and the conversion relationship to obtain the positioning result of the component defect in the photovoltaic power station yard.

9. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored on the memory and executable on the processor. When the program or instructions are executed by the processor, the steps of a method for locating component defects in a photovoltaic power station as described in any one of claims 1-7 are implemented.

10. A readable storage medium, characterized in that, A program or instructions are stored on the readable storage medium. When the program or instructions are executed by the processor, the steps of a method for locating component defects in a photovoltaic power station as described in any one of claims 1-7 are implemented.

Images