Abnormal identification method and system for photovoltaic junction box

Abstract

A method and system for identifying abnormalities of photovoltaic junction boxes, relating to the technical field of equipment monitoring. In this method, a preset area is divided to obtain a plurality of preset sub-areas; the area information corresponding to each preset sub-area is obtained; the detection images corresponding to all photovoltaic junction boxes in each preset sub-area are obtained; image recognition is performed on each detection image to obtain a first abnormal risk score corresponding to each photovoltaic junction box; each preset sub-area is identified based on each area information to obtain a second abnormal risk score corresponding to each preset sub-area; based on each first abnormal risk score and second abnormal risk score, it is determined that there are abnormal photovoltaic junction boxes in each preset sub-area. The implementation of the technical solution of the present application can effectively improve the accuracy of identifying abnormal working conditions of photovoltaic junction boxes.

Description

Technical Field

[0001] The present application relates to the technical field of equipment monitoring, and in particular to a method and system for identifying anomalies of a photovoltaic junction box. Background Art

[0002] With the development and application of photovoltaic power generation technology, photovoltaic junction boxes, as one of the indispensable components in solar photovoltaic power stations, play the role of connecting photovoltaic panels and inverters, protecting photovoltaic panels and inverters from external environmental factors. Therefore, in the process of photovoltaic power generation, it is necessary to monitor the working status of the photovoltaic junction box and accurately identify photovoltaic junction boxes with abnormal working status. At present, the monitoring of photovoltaic junction boxes is mainly carried out by staff to inspect the photovoltaic junction boxes, and the identification of photovoltaic junction boxes with abnormal working status depends on the work experience of the staff. For large-scale photovoltaic power stations, the accuracy of existing methods in identifying abnormal working status of photovoltaic junction boxes is low. Therefore, how to improve the accuracy of identifying abnormal working status of photovoltaic junction boxes has become a technical problem that needs to be solved urgently.

[0003] Therefore, there is an urgent need for an intelligent monitoring method and system for photovoltaic junction boxes to solve the problems existing in current technologies. Summary of the invention

[0004] The present application provides a method and system for identifying an abnormality of a photovoltaic junction box, which can effectively improve the accuracy of identifying an abnormal working state of a photovoltaic junction box.

[0005] In a first aspect, the present application provides a method for identifying abnormalities of photovoltaic junction boxes, the method comprising: dividing a preset area to obtain a plurality of preset sub-areas; obtaining area information corresponding to each of the preset sub-areas; obtaining detection images corresponding to all photovoltaic junction boxes in each of the preset sub-areas; performing image recognition on each of the detection images to obtain a first abnormal risk score corresponding to each of the photovoltaic junction boxes; identifying each of the preset sub-areas based on each of the area information to obtain a second abnormal risk score corresponding to each of the preset sub-areas; and determining, based on each of the first abnormal risk score and the second abnormal risk score, whether there is an abnormal photovoltaic junction box in each of the preset sub-areas.

[0006] By adopting the above technical solution, in order to facilitate the abnormal detection of photovoltaic junction boxes and improve the accuracy of detection, it is necessary to first divide the large area of ​​the preset area into multiple preset sub-areas with smaller areas. By obtaining the area information corresponding to each preset sub-area, the specific environment and maintenance history of each area can be considered, and more background information is introduced in the analysis and evaluation process, thereby enhancing the accuracy of risk assessment. By obtaining the detection images corresponding to all photovoltaic junction boxes in each preset sub-area, comprehensive coverage of each junction box is ensured. By performing image recognition on each detection image, the first abnormal risk score corresponding to each photovoltaic junction box is obtained, thereby identifying potential abnormal features. By identifying each preset sub-area based on each area information, the second abnormal risk score corresponding to each preset sub-area is obtained, thereby combining non-image data such as environmental factors and maintenance history, providing a more comprehensive risk assessment. By determining the abnormal photovoltaic junction boxes in each preset sub-area based on each first abnormal risk score and second abnormal risk score, a comprehensive assessment of the status of the photovoltaic junction box is achieved, thereby effectively improving the accuracy of identifying the abnormal working state of the photovoltaic junction box.

[0007] Optionally, the acquisition of detection images of all photovoltaic junction boxes in each of the preset sub-areas specifically includes: acquiring a first shooting distance between the current shooting device and the preset sub-area; acquiring a first initial image corresponding to the preset sub-area captured at the first shooting distance; performing weighted processing on the first initial image to obtain a first processed image; performing image enhancement processing on the first processed image to obtain a second processed image; performing preset contour detection on the second processed image to obtain a preset contour recognition result; the preset contour recognition result includes the number of contours corresponding to the second image and the contour area corresponding to each preset contour; when the number of contours meets the preset contour number, determining whether the areas of each contour are all within the area range corresponding to the first shooting distance; when the contour areas are all within the area range corresponding to the first shooting distance, obtaining each of the detection images based on the preset contour recognition result.

[0008] By adopting the above technical solution, by obtaining the first shooting distance of the current shooting device from the preset sub-area, it is ensured that the captured image can cover the entire sub-area, while maintaining sufficient image quality and detail clarity, so as to perform detailed image analysis and feature extraction. By performing weighted processing on the first initial image, a first processed image is obtained, thereby improving the visual effect of the image, highlighting important features, and facilitating more accurate subsequent image recognition and analysis. By performing image enhancement processing on the first processed image, a second processed image is obtained, thereby further improving the quality of the image, especially in terms of contrast, sharpness and details, making the image more suitable for detailed feature analysis. By performing preset contour detection on the second processed image, a preset contour recognition result is obtained, so that the photovoltaic junction box in the image can be accurately identified. When the number of contours meets the preset number of contours, it is determined whether the areas of each contour are all within the area range corresponding to the first shooting distance, thereby ensuring the accuracy and reliability of the image. When the areas of the contours are all within the area range corresponding to the first shooting distance, each detection image is obtained based on the preset contour recognition result, thereby ensuring that the final detection image contains accurate and complete photovoltaic junction box information.

[0009] Optionally, performing image recognition on each of the detection images to obtain a first abnormality risk score corresponding to each of the photovoltaic junction boxes specifically includes: based on each of the detection images, obtaining n image features and the number of image features corresponding to each of the detection images; wherein n is a positive integer greater than or equal to 1; based on the n image features, constructing an image feature vector I=[i1,…i m …,i n ]; when the number of image features corresponding to each of the detection images is greater than the preset number of features, the first abnormal risk score corresponding to each of the photovoltaic junction boxes is calculated based on each of the image feature vectors; when the number of image features corresponding to each of the detection images is not greater than the preset number of features, based on the first shooting distance, a distance correction factor is determined; based on the distance correction factor and each of the image feature vectors, the first abnormal risk score corresponding to each of the photovoltaic junction boxes is calculated.

[0010] By adopting the above technical solution, by obtaining n image features and the number of image features corresponding to each detection image based on each detection image, the key information related to the status of the photovoltaic junction box can be accurately captured, providing the necessary data basis for detailed risk analysis. By constructing the image feature vector corresponding to each detection image based on n image features, and when the number of image features corresponding to each detection image is greater than the preset number of features, the first abnormal risk score corresponding to each photovoltaic junction box is calculated based on each image feature vector, thereby ensuring that the calculation of the risk score is based on sufficient data and improving the accuracy and reliability of the score. When the number of image features corresponding to each detection image is not greater than the preset number of features, the distance correction factor is determined based on the first shooting distance, and the first abnormal risk score corresponding to each photovoltaic junction box is calculated based on the distance correction factor and each image feature vector, so that the scoring process can be adjusted to take into account the impact of the shooting distance on image quality and feature recognition to ensure the accuracy of the calculated first abnormal risk score.

[0011] Optionally, the regional information includes regional maintenance times and regional environmental factors; and the identifying of each of the preset sub-regions based on each of the regional information to obtain a second abnormal risk score corresponding to each of the preset sub-regions specifically includes: inputting the regional maintenance times into a preset risk assessment model to obtain a maintenance frequency score; inputting the regional environmental factors into a preset risk assessment model to obtain an environmental risk score; and calculating the second abnormal risk score based on the maintenance frequency score and the environmental risk score.

[0012] Optionally, the determining of the photovoltaic junction boxes with abnormalities in each of the preset sub-areas based on each of the first abnormal risk scores and the second abnormal risk score specifically includes: calculating the comprehensive abnormal risk score corresponding to each of the photovoltaic junction boxes based on each of the first abnormal risk scores and the second abnormal risk score; and screening the photovoltaic junction boxes with abnormalities from all the photovoltaic junction boxes based on each of the comprehensive abnormal risk scores; wherein the comprehensive abnormal risk score corresponding to the photovoltaic junction boxes with abnormalities is greater than a preset comprehensive abnormal risk score threshold.

[0013] Optionally, the calculating, based on each of the first abnormal risk scores and the second abnormal risk scores, a comprehensive abnormal risk score corresponding to each of the photovoltaic junction boxes specifically includes:

[0014] The comprehensive abnormal risk score is calculated by the following formula:

[0015] z=αf(x)+βy p ;

[0016]

[0017] Wherein, z is the comprehensive abnormal risk score; x is the first abnormal risk score; y is the second abnormal risk score; α is the first weight factor; β is the second weight factor; p is the non-linear adjustment coefficient, and p ∈ (0, 1); ε is the minimum threshold; k is the scaling coefficient, and k ∈ (0, 1); a is the piecewise limit constant, and a ∈ (0, 1).

[0018] By adopting the above technical solution, the first abnormal risk score x is adjusted by setting the piecewise function f(x). When x < a, the logarithmic function ln(max(x, ε)) + 1 is used to reduce the influence of smaller scores, and by setting a minimum threshold ε, the problem of the logarithmic function near zero is avoided. When x ≥ a, the exponential function e kx -1 is used to enhance the influence of higher scores, where k is the scaling coefficient and k ∈ (0, 1), which is used to control the growth rate of the exponential function. y p is the non-linear adjustment of the second abnormal risk score, where p is the non-linear adjustment coefficient. α is the first weight factor and β is the second weight factor, which are respectively used to adjust the influence of the two scores in the comprehensive score to ensure that the first abnormal risk score and the second abnormal risk score can reflect the actual risk situation.

[0019] Optionally, after performing the preset contour detection on the second processed image to obtain the preset contour recognition result, the method further includes: when the number of contours does not meet the preset number of contours, adjusting the first shooting distance to obtain a second shooting distance; acquiring a second initial image corresponding to the preset sub-region shot at the second shooting distance; performing weighted processing on the second initial image to obtain a third processed image; performing image enhancement processing on the third processed image to obtain a fourth processed image; performing preset contour detection on the fourth processed image to obtain the number of contours corresponding to the fourth image; if the number of contours still does not meet the preset number of contours, readjusting the second shooting distance until the number of contours meets the preset number of contours.

[0020] In a second aspect of the present application, a photovoltaic junction box abnormality identification system is provided, the system comprising an area division module, an acquisition module and a processing module; the area division module is used to divide a preset area to obtain a plurality of preset sub-areas; the acquisition module is used to obtain area information corresponding to each of the preset sub-areas; the acquisition module is also used to obtain detection images corresponding to all photovoltaic junction boxes in each of the preset sub-areas; the processing module is used to perform image recognition on each of the detection images to obtain a first abnormality risk score corresponding to each of the photovoltaic junction boxes; the processing module is also used to identify each of the preset sub-areas based on each of the area information to obtain a second abnormality risk score corresponding to each of the preset sub-areas; the processing module is also used to determine that there are abnormal photovoltaic junction boxes in each of the preset sub-areas based on each of the first abnormality risk scores and the second abnormality risk scores.

[0021] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes any method as described in the first aspect of the present application.

[0022] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program that can be loaded by a processor and execute the method as described in any one of the first aspects of the present application.

[0023] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0024] 1. To facilitate the anomaly detection of photovoltaic junction boxes and improve the accuracy of detection, it is necessary to first divide a large preset area into multiple smaller preset sub-areas. By obtaining the area information corresponding to each preset sub-area, it is possible to consider the specific environment and maintenance history of each area, introduce more background information in the analysis and evaluation process, and enhance the accuracy of risk assessment. By obtaining the detection images corresponding to all photovoltaic junction boxes within each preset sub-area, the comprehensive coverage of each junction box is ensured. By performing image recognition on each detection image, the first anomaly risk score corresponding to each photovoltaic junction box is obtained, thereby identifying potential anomaly features. By identifying each preset sub-area based on each area information, the second anomaly risk score corresponding to each preset sub-area is obtained, thereby combining non-image data such as environmental factors and maintenance history and providing a more comprehensive risk assessment. By determining the photovoltaic junction boxes with anomalies within each preset sub-area based on each first anomaly risk score and second anomaly risk score, the comprehensive assessment of the status of photovoltaic junction boxes is achieved, and thus the accuracy of identifying the abnormal working state of photovoltaic junction boxes can be effectively improved.

[0025] 2. By obtaining n image features and the number of image features corresponding to each detection image based on each detection image, the key information related to the status of photovoltaic junction boxes can be accurately captured, providing a necessary data basis for detailed risk analysis. By constructing the image feature vector corresponding to each detection image based on the n image features, and when the number of image features corresponding to each detection image is greater than the preset number of features, calculating the first anomaly risk score corresponding to each photovoltaic junction box based on each image feature vector, the calculation of the risk score is ensured to be based on sufficient data, improving the accuracy and reliability of the score. When the number of image features corresponding to each detection image is not greater than the preset number of features, determining the distance correction factor based on the first shooting distance, and calculating the first anomaly risk score corresponding to each photovoltaic junction box based on the distance correction factor and each image feature vector, the scoring process can be adjusted to consider the influence of the shooting distance on image quality and feature recognition to ensure the accuracy of the calculated first anomaly risk score.

[0026] 3. By setting a piecewise function f(x) to adjust the first anomaly risk score x. When x < a, the logarithmic function ln(max(x, ε)) + 1 is used to reduce the influence of smaller scores, and by setting a minimum threshold ε, the problem of the logarithmic function near zero is avoided. When x ≥ a, the exponential function e kx -1 is used to enhance the influence of higher scores, where k is a scaling factor and k ∈ (0, 1), used to control the growth rate of the exponential function. y pIt is a nonlinear adjustment of the second abnormal risk score, where p is the nonlinear adjustment coefficient. α is the first weight factor and β is the second weight factor, which are used to adjust the impact of the two scores in the comprehensive score to ensure that the first abnormal risk score and the second abnormal risk score can reflect the actual risk situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of a method for identifying an abnormality of a photovoltaic junction box provided in an embodiment of the present application;

[0028] Figure 2 It is a structural schematic diagram of an abnormality identification system for a photovoltaic junction box provided in an embodiment of the present application;

[0029] Figure 3 It is a structural schematic diagram of an electronic device disclosed in an embodiment of the present application.

[0030] Explanation of the reference numerals: 1. Area division module; 2. Acquisition module; 3. Processing module; 300. Electronic device; 301. Processor; 302. Communication bus; 303. User interface; 304. Network interface; 305. Memory. DETAILED DESCRIPTION

[0031] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0032] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0033] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" can explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0034] This application provides a method for identifying abnormalities in a photovoltaic junction box. Figure 1 , which shows a schematic flow diagram of a method for identifying abnormalities of a photovoltaic junction box provided by an embodiment of the present application. The method includes steps S1-S6, which are as follows: Step S1: Divide a preset area to obtain multiple preset sub-areas.

[0035] Specifically, in the present technical solution, the preset area is a large-area photovoltaic power station. Since the preset area is large, in order to facilitate the abnormal detection of the photovoltaic junction box and improve the accuracy of the detection, it is necessary to first divide the preset area into multiple smaller preset sub-areas.

[0036] When dividing the preset areas, the server needs to consider several factors such as the physical layout of the junction box, lighting conditions, and maintenance convenience. The size and shape of multiple preset sub-areas should be determined according to actual conditions, which should facilitate image acquisition and ensure effective monitoring.

[0037] Step S2: Obtaining the area information corresponding to each preset sub-area.

[0038] Specifically, in this technical solution, the regional information includes the regional maintenance times and regional environmental factors. The server will collect the historical maintenance data of the photovoltaic junction boxes in each preset sub-region to obtain the regional maintenance times. The regional environmental factors are obtained through environmental monitoring equipment, including but not limited to the climate conditions (such as temperature, humidity, wind speed), geographical features (such as orientation, inclination), etc. of the region.

[0039] Step S3: Acquire detection images corresponding to all photovoltaic junction boxes in each preset sub-area.

[0040] Specifically, in the present technical solution, the server obtains detection images corresponding to all photovoltaic junction boxes in each preset sub-area.

[0041] First, an initial image is captured by a shooting device, and the server then performs a series of processing on the initial image to obtain a detection image corresponding to all photovoltaic junction boxes in each preset sub-area. The shooting device may be a shooting drone or a camera. In this solution, the shooting device is preferably a shooting drone.

[0042] The following embodiment will describe in detail the specific method of obtaining the detection images corresponding to all photovoltaic junction boxes in each preset sub-area.

[0043] In a possible implementation, step S3 specifically includes the following steps:

[0044] Obtain a first shooting distance between the current shooting device and a preset sub-area.

[0045] Specifically, in the present technical solution, the server obtains a first shooting distance between the current shooting device and a preset sub-area, and the first shooting distance may be a uniformly varying distance such as 10m, 15m, or 20m.

[0046] A first initial image corresponding to a preset sub-region captured at a first shooting distance is acquired.

[0047] Specifically, in the present technical solution, the server will send a control signal to the current shooting device to control the current shooting device to shoot at a first shooting distance to obtain a first initial image corresponding to a preset sub-area.

[0048] The first initial image is weighted to obtain a first processed image.

[0049] Specifically, in this technical solution, the server performs weighted processing on the first initial image to obtain a first processed image. The weighted processing is intended to improve the image quality of the first initial image so that the features of all photovoltaic junction boxes extracted subsequently are clearer.

[0050] To perform weighted processing on the first initial image, it is necessary to first perform smoothing filtering processing on each first initial image to obtain an adjusted image corresponding to each first initial image; then determine whether the SSIM value corresponding to the adjusted image is greater than a preset value; if the SSIM value corresponding to the adjusted image is greater than the preset value, perform gradient histogram calculation on each adjusted image to obtain multiple regions of interest, and determine whether the feature information of the photovoltaic junction box in each region of interest is empty; if the feature information of the photovoltaic junction box in each region of interest is not empty, compare each region of interest with the pixels of the first initial image corresponding to the region of interest to obtain pixel factors of multiple regions of interest; then obtain the true value of the pixel of each region of interest; and based on the pixel factor and the true value of the pixel, calculate the weighted pixel values ​​of the multiple regions of interest; finally, based on the weighted pixel values ​​of the multiple regions of interest, correspondingly weight the pixels of each region of interest to obtain multiple weighted first processed images.

[0051] More specifically, each first initial image is subjected to a smoothing filter process to remove some noise in the first initial image and make the first initial image smoother. Commonly used smoothing filters include Gaussian filters and mean filters.

[0052] When the server determines that the SSIM value corresponding to the adjusted image is greater than the preset value, the Sobel operator or Scharr operator is used to calculate the horizontal and vertical gradients of the adjusted image, and the amplitudes of the horizontal and vertical gradients are merged into a gradient image, that is, the modulus of the gradient is calculated at each pixel position. Then the gradient image is histogram calculated, and the region of interest is determined according to the peak value or threshold of the histogram. If the server determines that the SSIM value corresponding to the adjusted image is not greater than the preset value, the parameters of the smoothing filter processing need to be adjusted, and each first initial image is smoothed again.

[0053] When the server determines that the characteristic information of the photovoltaic module in each region of interest is not empty, each region of interest is compared with the pixel of the distance image corresponding to the region of interest, and the pixel factors of multiple regions of interest are obtained. The average value, median value or other statistical features of the pixel values ​​in the region can be calculated for each region of interest. And in the corresponding distance image, the average value, median value or other statistical features of the pixel values ​​of the distance image are calculated. The ratio of the two is the pixel factor of the region of interest.

[0054] The weighted pixel value of the region of interest is then calculated using the following formula: weighted pixel value = 1 / (1-pixel factor) × true pixel value.

[0055] Finally, by correspondingly weighting the pixel value of each region of interest, for example, weighting different target regions to highlight their importance in the entire image, a plurality of weighted first processed images are obtained.

[0056] Perform image enhancement processing on the first processed image to obtain a second processed image.

[0057] Specifically, in this technical solution, the server performs image enhancement processing on each weighted first processed image to obtain multiple second processed images after image enhancement. That is, HSV processing is performed on each weighted first processed image. HSV processing refers to color space conversion of pixels in an image or video, converting RGB (red, green, and blue) color space into HSV (hue, saturation, and brightness) color space.

[0058] Perform preset contour detection on the second processed image to obtain preset contour recognition results; the preset contour recognition results include the number of contours corresponding to the second image and contour areas corresponding to each preset contour.

[0059] Specifically, in the present technical solution, the preset contour detection is performed on the second processed image, that is, the second processed image is identified according to the preset contour of the photovoltaic junction box, and the number of contours corresponding to the second image and the contour area corresponding to each preset contour are obtained. The method for detecting the preset contour frame can be to use the contour detection algorithm in image processing (such as the findContours function in OpenCV) to detect all contours in the image. And some contours that do not meet the requirements are excluded through some screening conditions. For the retained contours, a rectangular fitting algorithm (such as the minAreaRect function in OpenCV) is used to fit the contour area corresponding to the preset contour. And the number of contours corresponding to the second image is obtained.

[0060] When the number of contours meets the preset number of contours, it is determined whether the areas of the contours are all within the area range corresponding to the first shooting distance.

[0061] Specifically, in the present technical solution, before determining whether each contour area is within the area range corresponding to the first shooting distance, the experimenter first needs to shoot the photovoltaic junction box in the preset sub-area multiple times at each different shooting distance, and count the area of ​​the photovoltaic junction box at each shooting distance to obtain the area range corresponding to the first shooting distance.

[0062] When the number of contours meets the preset number of contours, that is, when the number of identified photovoltaic junction boxes meets the number of photovoltaic junction boxes in the actual preset sub-area, the server will determine whether the areas of each contour are within the area range corresponding to the first shooting distance.

[0063] When the contour areas are all within the area range corresponding to the first shooting distance, each detection image is obtained based on the preset contour recognition result.

[0064] Specifically, in the present technical solution, if the server determines that the area of ​​each contour is within the area range corresponding to the current distance, the recognition image obtained based on the preset contour recognition result is used as each detection image.

[0065] Step S4: performing image recognition on each detection image to obtain a first abnormal risk score corresponding to each photovoltaic junction box.

[0066] Specifically, in the present technical solution, the server performs image recognition on each detection image to obtain a first abnormal risk score corresponding to each photovoltaic junction box.

[0067] In a possible implementation, step S4 specifically includes the following steps:

[0068] Based on each detection image, n image features and the number of image features corresponding to each detection image are obtained; wherein n is a positive integer greater than or equal to 1.

[0069] Specifically, in this technical solution, the server will analyze each detection image and extract n key image features from each detection image. These image features need to be able to accurately reflect the appearance of the photovoltaic junction box. More specifically, the image features include edge features, texture features, and shape features.

[0070] Based on n image features, construct the image feature vector I=[i1,…i m …,i n ].

[0071] Specifically, in this technical solution, the server integrates the extracted n image features to construct an image feature vector I = [i1, ...i m …,i n ]. The image feature vector will be used for the subsequent calculation of the first abnormality risk score.

[0072] When the number of image features corresponding to each detection image is greater than a preset number of features, a first abnormality risk score corresponding to each photovoltaic junction box is calculated based on each image feature vector.

[0073] Specifically, in this technical solution, when the number of image features corresponding to each detection image is greater than the preset number of features, it means that the detection image obtained at the current first shooting distance is clear enough and the image quality is good, and a large number of image features can be extracted. Therefore, the n image features in each image feature vector are converted into feature values ​​through the image feature scoring function, so as to calculate the first abnormal risk score corresponding to each photovoltaic junction box. The first abnormal risk score can be expressed by the following formula:

[0074]

[0075] Among them, E1 is the first abnormal risk score, i m is the mth image feature, f(i m ) is the mth image feature scoring function. The image feature scoring function is used to convert image feature i m Convert to a standardized risk score. Image feature scoring function f(i m ) can be a statistical model or a machine learning algorithm.

[0076] When the number of image features corresponding to each detection image is not greater than the preset number of features, a distance correction factor is determined based on the first shooting distance.

[0077] Specifically, in this technical solution, when the number of image features corresponding to each detection image is not greater than the preset number of features, it means that the detection image obtained at the current first shooting distance is not clear enough and the image quality is poor. Therefore, it is necessary to determine the distance correction factor based on the first shooting distance to correct the detectable image features so that the image features are converted into standardized risk scores that are more in line with the actual working conditions of the photovoltaic junction box. Different shooting distances correspond to different distance correction factors. The distance correction factor can be determined based on historical data analysis or expert knowledge. Therefore, the value of the distance correction factor is not specifically limited in this solution.

[0078] Based on the distance correction factor and each image feature vector, a first abnormal risk score corresponding to each photovoltaic junction box is calculated.

[0079] Specifically, in this technical solution, when the number of image features corresponding to each detection image is not greater than the preset number of features, the server calculates the first abnormal risk score corresponding to each photovoltaic junction box based on the distance correction factor and each image feature vector. At this time, the first abnormal risk score can be expressed by the following formula:

[0080]

[0081] Among them, x is the first abnormal risk score, i m is the mth image feature, f(i m ) is the mth image feature scoring function, w d is the distance correction factor. The image feature scoring function is used to convert the image feature i m Convert to a standardized risk score. Image feature scoring function f(i m ) can be a statistical model or a machine learning algorithm.

[0082] Step S5: identifying each preset sub-region based on each region information, and obtaining a second abnormal risk score corresponding to each preset sub-region.

[0083] Specifically, in this technical solution, since the area information includes the area maintenance times and the area environmental factors, the server will identify each preset sub-area based on the area maintenance times and the area environmental factors to obtain the second abnormal risk score corresponding to each preset sub-area.

[0084] In a possible implementation, step S5 specifically includes the following steps:

[0085] The regional maintenance times are input into the preset risk assessment model to obtain the maintenance frequency score.

[0086] Specifically, in this technical solution, the server inputs the number of regional maintenance times into a preset risk assessment model. The preset risk assessment model can be based on statistical analysis, historical data comparison or machine learning algorithm to evaluate the relationship between maintenance frequency and potential abnormal risks, thereby outputting a maintenance frequency score.

[0087] Input regional environmental factors into the preset risk assessment model to obtain an environmental risk score.

[0088] Specifically, in the present technical solution, regional environmental factors include but are not limited to temperature, humidity, wind speed, sunshine intensity, etc.; the server inputs the regional maintenance times into a preset risk assessment model, which can be based on statistical analysis, historical data comparison or machine learning algorithms to evaluate the relationship between regional environmental factors and potential abnormal risks, thereby outputting an environmental risk score.

[0089] Based on the maintenance frequency score and the environmental risk score, a second abnormal risk score is calculated.

[0090] Specifically, in this technical solution, the second abnormal risk score can be expressed by the following formula:

[0091] y=μk1+θk2;

[0092] Wherein, y is the first abnormal risk score, k1 is the maintenance frequency score, k2 is the environmental risk score, μ is the first impact weight, θ is the second impact weight, and μ+θ=1. The first impact weight and the second impact weight can be determined based on historical data analysis or expert knowledge. Therefore, the values ​​of the first impact weight and the second impact weight are not specifically limited in this solution.

[0093] Step S6: Based on each of the first abnormality risk scores and the second abnormality risk scores, determining whether there is an abnormal photovoltaic junction box in each of the preset sub-areas.

[0094] Specifically, in the present technical solution, the server determines whether there are abnormal photovoltaic junction boxes in each preset sub-area based on each first abnormal risk score and second abnormal risk score.

[0095] In a possible implementation, step S6 specifically includes the following steps:

[0096] Based on each first abnormal risk score and the second abnormal risk score, a comprehensive abnormal risk score corresponding to each photovoltaic junction box is calculated. Based on each comprehensive abnormal risk score, photovoltaic junction boxes with abnormalities are screened from all photovoltaic junction boxes; wherein the comprehensive abnormal risk score corresponding to the photovoltaic junction box with abnormalities is greater than a preset comprehensive abnormal risk score threshold.

[0097] Specifically, in this technical solution, the server first calculates the comprehensive abnormal risk score corresponding to each photovoltaic junction box based on each first abnormal risk score and the second abnormal risk score. The server then determines a preset comprehensive abnormal risk score threshold, which can be determined based on factors such as historical data analysis, equipment performance standards, and risk acceptance. When the comprehensive abnormal risk score corresponding to the photovoltaic junction box is greater than the preset comprehensive abnormal risk score threshold, the server can determine that the photovoltaic junction box is an abnormal photovoltaic junction box.

[0098] In a possible implementation manner, the comprehensive abnormal risk score is calculated through the following formula:

[0099] z = αf(x) + βy p ;

[0100]

[0101] where z is the comprehensive abnormal risk score; x is the first abnormal risk score; y is the second abnormal risk score; α is the first weight factor; β is the second weight factor; p is the nonlinear adjustment coefficient, and p ∈ (0, 1); ε is the minimum threshold; k is the scaling coefficient, and k ∈ (0, 1); a is the piecewise limit constant, and a ∈ (0, 1).

[0102] Specifically, in this technical solution, f(x) is a piecewise function used to adjust the first abnormal risk score x. When x < a, the logarithmic function ln(max(x, ε)) + 1 is used to reduce the influence of smaller scores. Among them, by setting a minimum threshold ε, ε is preferably 0.01 to avoid problems of the logarithmic function near zero. When x ≥ a, the exponential function e kx -1 is used to enhance the influence of higher scores, where k is the scaling coefficient and k ∈ (0, 1), which is used to control the growth rate of the exponential function.

[0103] y p is the nonlinear adjustment of the second abnormal risk score, where p is the nonlinear adjustment coefficient.

[0104] α is the first weight factor, and β is the second weight factor, which are respectively used to adjust the influence of the two scores in the comprehensive score. The first weight factor and the second weight factor need to be based on actual data analysis or expert opinions to ensure that the first abnormal risk score and the second abnormal risk score can reflect the actual risk situation.

[0105] In a possible implementation manner, after performing preset contour detection on the second processed image to obtain a preset contour recognition result, the method further includes the following steps:

[0106] When the number of contours does not meet the preset number of contours, the first shooting distance is adjusted to obtain a second shooting distance.

[0107] Specifically, in the present technical solution, when the number of contours does not meet the preset number of contours, it indicates that the shooting distance is too far or too close, resulting in that the photovoltaic junction boxes in the image cannot be clearly distinguished or part of the junction boxes are not captured. If the shooting distance is too far, the image resolution may be insufficient, and the outline of the photovoltaic junction box may not be obvious or too small. In this case, the shooting distance needs to be reduced. Conversely, if the shooting distance is too close, some photovoltaic junction boxes may not be completely captured. At this time, the shooting distance should be increased. The first shooting distance is adjusted to obtain a distance adjustment method for the second shooting distance. The adjustment can be based on empirical rules (for example, increasing or decreasing a fixed percentage) or based on more precise calculations (taking into account factors such as the focal length of the camera and the size of the sensor).

[0108] A second initial image corresponding to the preset sub-region captured at a second shooting distance is acquired.

[0109] Specifically, in the present technical solution, the method for obtaining the second initial image corresponding to the preset sub-area captured at the second shooting distance is similar to the method for obtaining the first initial image corresponding to the preset sub-area captured at the first shooting distance in the aforementioned embodiment, so no further details will be given here.

[0110] The second initial image is weightedly processed to obtain a third processed image.

[0111] Specifically, in the present technical solution, the method of weighted processing the second initial image to obtain the third processed image is similar to the method of weighted processing the first initial image to obtain the first processed image in the aforementioned embodiment, so no further details will be given here.

[0112] Perform image enhancement processing on the third processed image to obtain a fourth processed image.

[0113] Specifically, in the present technical solution, the method of performing image enhancement processing on the third processed image to obtain the fourth processed image is similar to the method of performing image enhancement processing on the first processed image to obtain the second processed image in the aforementioned embodiment, so no further details will be given here.

[0114] Perform preset contour detection on the fourth processed image to obtain the number of contours corresponding to the fourth image.

[0115] Specifically, in the present technical solution, the method of performing preset contour detection on the fourth processed image to obtain the number of contours corresponding to the fourth image is similar to the method of performing preset contour detection on the second processed image to obtain the number of contours corresponding to the second image in the aforementioned embodiment, so no further details will be given here.

[0116] If the number of contours still does not meet the preset number of contours, the second shooting distance is adjusted again until the number of contours meets the preset number of contours.

[0117] Specifically, in this technical solution, if the server determines that the number of contours still does not meet the preset number of contours, the second shooting distance needs to be adjusted again until the number of contours meets the preset number of contours. When the number of contours meets the preset number of contours, the server performs the subsequent step of determining whether the areas of each contour are within the area range corresponding to the first shooting distance.

[0118] Reference Figure 2 , which shows a structural schematic diagram of an abnormality identification system for a photovoltaic junction box provided in an embodiment of the present application, the system includes an area division module 1, an acquisition module 2 and a processing module 3; the area division module 1 is used to divide a preset area to obtain multiple preset sub-areas; the acquisition module 2 is used to obtain area information corresponding to each preset sub-area; the acquisition module 2 is also used to obtain detection images corresponding to all photovoltaic junction boxes in each preset sub-area; the processing module 3 is used to perform image recognition on each detection image to obtain a first abnormality risk score corresponding to each photovoltaic junction box; the processing module 3 is also used to identify each preset sub-area based on each area information to obtain a second abnormality risk score corresponding to each preset sub-area; the processing module 3 is also used to determine that there are abnormal photovoltaic junction boxes in each preset sub-area based on each first abnormality risk score and the second abnormality risk score.

[0119] In a possible implementation, the acquisition module 2 is also used to acquire a first shooting distance between the current shooting device and the preset sub-area; the acquisition module 2 is also used to acquire a first initial image corresponding to the preset sub-area captured at the first shooting distance; the processing module 3 is also used to perform weighted processing on the first initial image to obtain a first processed image; the processing module 3 is also used to perform image enhancement processing on the first processed image to obtain a second processed image; the processing module 3 is also used to perform preset contour detection on the second processed image to obtain a preset contour recognition result; the preset contour recognition result includes the number of contours corresponding to the second image and the contour area corresponding to each preset contour; the processing module 3 is also used to determine whether each contour area is within the area range corresponding to the first shooting distance when the number of contours meets the preset number of contours; the processing module 3 is also used to obtain each detection image based on the preset contour recognition result when the contour areas are all within the area range corresponding to the first shooting distance.

[0120] In a possible implementation, the processing module 3 is further used to obtain n image features and the number of image features corresponding to each detection image based on each detection image; wherein n is a positive integer greater than or equal to 1; the processing module 3 is further used to construct an image feature vector I=[i1,…i m …,i n ]; the processing module 3 is also used to calculate the first abnormal risk score corresponding to each photovoltaic junction box based on each image feature vector when the number of image features corresponding to each detection image is greater than the preset number of features; the processing module 3 is also used to determine the distance correction factor based on the first shooting distance when the number of image features corresponding to each detection image is not greater than the preset number of features; the processing module 3 is also used to calculate the first abnormal risk score corresponding to each photovoltaic junction box based on the distance correction factor and each image feature vector.

[0121] In one possible implementation, the processing module 3 is also used to input the number of regional maintenance times into a preset risk assessment model to obtain a maintenance frequency score; the processing module 3 is also used to input regional environmental factors into a preset risk assessment model to obtain an environmental risk score; the processing module 3 is also used to calculate a second abnormal risk score based on the maintenance frequency score and the environmental risk score.

[0122] In a possible implementation, the processing module 3 is further used to calculate the comprehensive abnormal risk score corresponding to each photovoltaic junction box based on each first abnormal risk score and the second abnormal risk score; the processing module 3 is further used to screen out abnormal photovoltaic junction boxes from all photovoltaic junction boxes based on each comprehensive abnormal risk score; wherein the comprehensive abnormal risk score corresponding to the abnormal photovoltaic junction box is greater than a preset comprehensive abnormal risk score threshold.

[0123] In a possible implementation, the processing module 3 is further configured to calculate a comprehensive abnormal risk score using the following formula:

[0124] z=αf(x)+βy p ;

[0125]

[0126] Among them, z is the comprehensive abnormal risk score; x is the first abnormal risk score; y is the second abnormal risk score; α is the first weight factor; β is the second weight factor; p is the nonlinear adjustment coefficient, and p∈(0,1); ε is the minimum threshold; k is the scaling factor, and k∈(0,1); a is the piecewise limit constant, and a∈(0,1).

[0127] In a possible implementation, the processing module 3 is further used to adjust the first shooting distance to obtain a second shooting distance when the number of contours does not meet the preset number of contours; the acquisition module 2 is further used to obtain a second initial image corresponding to a preset sub-area captured at the second shooting distance; the processing module 3 is further used to perform weighted processing on the second initial image to obtain a third processed image; the processing module 3 is further used to perform image enhancement processing on the third processed image to obtain a fourth processed image; the processing module 3 is further used to perform preset contour detection on the fourth processed image to obtain the number of contours corresponding to the fourth image; the processing module 3 is further used to adjust the second shooting distance again when the number of contours still does not meet the preset number of contours until the number of contours meets the preset number of contours.

[0128] It should be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0129] The present application also discloses an electronic device. Figure 3 , Figure 3 The electronic device 300 may include: at least one processor 301 , at least one network interface 304 , a user interface 303 , a memory 305 , and at least one communication bus 302 .

[0130] The communication bus 302 is used to realize the connection and communication between these components.

[0131] The user interface 303 may include a display screen (Display) and a camera (Camera). The optional user interface 303 may also include a standard wired interface and a wireless interface.

[0132] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0133] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or more combinations of a central processing unit (Central Processing Unit, CPU), a graphics processor (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented by a single chip.

[0134] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally be at least one storage device located away from the aforementioned processor 301. Refer to Figure 3 , the memory 305 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module and an application program.

[0135] exist Figure 3In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call an application stored in the memory 305, and when executed by one or more processors 301, the electronic device 300 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0136] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0137] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0138] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0139] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0140] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0141] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0142] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for identifying abnormalities of a photovoltaic junction box, characterized in that: The method comprises: Dividing the preset area to obtain a plurality of preset sub-areas; Obtaining area information corresponding to each of the preset sub-areas; Acquire a first shooting distance between the current shooting device and the preset sub-area; Acquire a first initial image corresponding to the preset sub-area captured at the first shooting distance; Performing weighted processing on the first initial image to obtain a first processed image; Performing image enhancement processing on the first processed image to obtain a second processed image; Performing preset contour detection on the second processed image to obtain preset contour recognition results; the preset contour recognition results include the number of contours corresponding to the second processed image and contour areas corresponding to each preset contour; When the number of contours meets the preset number of contours, determining whether the areas of the contours are all within the area range corresponding to the first shooting distance; When the contour areas are all within the area range corresponding to the first shooting distance, obtaining each detection image based on the preset contour recognition result; Performing image recognition on each of the detection images to obtain a first abnormality risk score corresponding to each of the photovoltaic junction boxes; Identify each of the preset sub-regions based on the regional information to obtain a second abnormality risk score corresponding to each of the preset sub-regions; Based on each of the first abnormality risk scores and the second abnormality risk scores, it is determined that an abnormal photovoltaic junction box exists in each of the preset sub-areas.

2. The method according to claim 1, characterized in that The performing image recognition on each of the detection images to obtain a first abnormality risk score corresponding to each of the photovoltaic junction boxes specifically includes: Based on each of the detection images, obtain n image features and the number of image features corresponding to each of the detection images; wherein n is a positive integer greater than or equal to 1; Based on the n image features, an image feature vector I=[i1, ..., i m , ..., i n ]; When the number of image features corresponding to each of the detection images is greater than a preset number of features, the first abnormality risk score corresponding to each of the photovoltaic junction boxes is calculated based on each of the image feature vectors; When the number of the image features corresponding to each of the detection images is not greater than a preset number of features, determining a distance correction factor based on the first shooting distance; The first abnormality risk score corresponding to each photovoltaic junction box is calculated based on the distance correction factor and each of the image feature vectors.

3. The method according to claim 1, characterized in that The regional information includes the number of regional maintenance times and regional environmental factors; the identifying each of the preset sub-regions based on each of the regional information to obtain the second abnormal risk score corresponding to each of the preset sub-regions specifically includes: Inputting the maintenance frequency of the area into a preset risk assessment model to obtain a maintenance frequency score; Inputting the regional environmental factors into a preset risk assessment model to obtain an environmental risk score; The second abnormal risk score is calculated based on the maintenance frequency score and the environmental risk score.

4. The method according to claim 1, characterized in that: The determining, based on each of the first abnormality risk score and the second abnormality risk score, that there is an abnormal photovoltaic junction box in each of the preset sub-areas specifically includes: Calculating a comprehensive abnormality risk score corresponding to each of the photovoltaic junction boxes based on each of the first abnormality risk scores and the second abnormality risk score; Based on each of the comprehensive abnormality risk scores, the photovoltaic junction boxes with abnormalities are screened out from all the photovoltaic junction boxes; wherein the comprehensive abnormality risk score corresponding to the photovoltaic junction box with abnormality is greater than a preset comprehensive abnormality risk score threshold.

5. The method according to claim 1, characterized in that: The calculating, based on each of the first abnormal risk scores and the second abnormal risk scores, a comprehensive abnormal risk score corresponding to each of the photovoltaic junction boxes specifically includes: The comprehensive abnormal risk score is calculated by the following formula: z=αf(x)+βy p ; Among them, z is the comprehensive abnormal risk score; x is the first abnormal risk score; y is the second abnormal risk score; α is the first weight factor; β is the second weight factor; p is the nonlinear adjustment coefficient, and p∈(0,1); ε is the minimum threshold; k is the scaling factor, and k∈(0,1); a is the piecewise limit constant, and a∈(0,1).

6. The method according to claim 1, characterized in that After performing preset contour detection on the second processed image to obtain a preset contour recognition result, the method further includes: When the number of contours does not meet the preset number of contours, the first shooting distance is adjusted to obtain a second shooting distance; and a second initial image corresponding to the preset sub-area obtained by shooting at the second shooting distance is acquired; performing weighted processing on the second initial image to obtain a third processed image; Performing image enhancement processing on the third processed image to obtain a fourth processed image; Performing preset contour detection on the fourth processed image to obtain the number of contours corresponding to the fourth processed image; If the number of contours still does not meet the preset number of contours, the second shooting distance is adjusted again until the number of contours meets the preset number of contours.

7. A photovoltaic junction box abnormality identification system for executing the photovoltaic junction box abnormality identification method according to claim 1, characterized in that: The system includes a region division module, an acquisition module and a processing module; The region division module is used to divide the preset region into multiple preset sub-regions; The acquisition module is used to acquire the area information corresponding to each of the preset sub-areas; The acquisition module is further used to acquire detection images corresponding to all photovoltaic junction boxes in each of the preset sub-areas; The processing module is used to perform image recognition on each of the detection images to obtain a first abnormality risk score corresponding to each of the photovoltaic junction boxes; The processing module is further used to identify each of the preset sub-regions based on the information of each region, and obtain a second abnormal risk score corresponding to each of the preset sub-regions; The processing module is further configured to determine, based on each of the first abnormality risk scores and the second abnormality risk scores, whether there is an abnormal photovoltaic junction box in each of the preset sub-areas.

8. An electronic device, characterized in that: The electronic device (300) comprises a processor (301), a memory (305), a user interface (303) and a network interface (304), wherein the memory (305) is used to store instructions, the user interface (303) and the network interface (304) are used to communicate with other devices, and the processor (301) is used to execute the instructions stored in the memory (305) so that the electronic device (300) executes the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 6 is performed.

Images