Abnormal positioning method and system for photovoltaic module

Abstract

The invention relates to the technical field of photovoltaic power generation, and particularly discloses a photovoltaic module abnormity positioning method and system, and the method comprises the steps: obtaining the infrared image data of a photovoltaic module, segmenting the infrared image data of the photovoltaic module, and obtaining a plurality of photovoltaic module segmented images; calculating an abnormal coefficient of the segmented image of the photovoltaic module, and screening out an abnormal image of the photovoltaic module according to the abnormal coefficient of the segmented image of the photovoltaic module; acquiring a feature vector of the abnormal image of the photovoltaic module, and determining an abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; and determining an early warning weight according to the abnormal type of the abnormal image of the photovoltaic module, performing early warning on the abnormal image of the photovoltaic module according to the early warning weight and the abnormal coefficient, and positioning the corresponding photovoltaic module. According to the invention, the abnormal photovoltaic module image is segmented through the abnormal coefficient and the early warning weight, and early warning positioning is carried out, so that the abnormal photovoltaic area can be positioned in time, and the maintenance efficiency of the photovoltaic module is improved.

Description

Technical Field

[0001] The present application relates to the technical field of photovoltaic power generation, and more specifically, to a method and system for abnormal positioning of photovoltaic modules. Background Art

[0002] As the core part of a photovoltaic power generation system, the stability and reliability of photovoltaic modules directly affect the power generation efficiency and operation safety of the entire system. However, during actual operation, photovoltaic modules may fail due to various reasons. How to quickly and accurately locate and solve these failures has become an important task for maintenance personnel.

[0003] When using an infrared thermal imaging lens to detect abnormalities in photovoltaic modules in the prior art, it is difficult to detect the specific failure types and severity levels of photovoltaic modules, and the accuracy of abnormal detection is low, resulting in frequent false alarms and low maintenance efficiency in power plants. Summary of the Invention

[0004] The present invention provides a method and system for abnormal positioning of photovoltaic modules to solve the problem of low accuracy of abnormal detection of photovoltaic modules in the prior art, including: Obtain infrared image data of a photovoltaic module, segment the infrared image data of the photovoltaic module to obtain a plurality of segmented images of the photovoltaic module; Calculate the abnormality coefficient of the segmented image of the photovoltaic module, and screen out the abnormal images of the photovoltaic module according to the abnormality coefficient of the segmented image of the photovoltaic module; Obtain the feature vector of the abnormal image of the photovoltaic module, and determine the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; Determine the warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and perform warning on the abnormal image of the photovoltaic module and locate the corresponding photovoltaic module according to the warning weight and the abnormality coefficient.

[0005] Further, the segmenting the infrared image data of the photovoltaic module to obtain a plurality of segmented images of the photovoltaic module includes: Perform grayscale processing on the infrared image data of the photovoltaic module to obtain a grayscale image of the photovoltaic module; Perform edge detection on the grayscale image of the photovoltaic module based on the canny edge detection algorithm, and determine the fault edge according to the edge detection result of the grayscale image of the photovoltaic module; Obtain a preset standard fault edge, and calculate the fault fitting degree between the fault edge of the current grayscale image of the photovoltaic module and the preset standard fault edge; Cluster each pixel point according to the fault fitting degree and the grayscale value of the pixel point located on the fault edge, and connect the adjacent pixel points in the same clustering cluster according to the clustering result to obtain a segmented image of the photovoltaic module.

[0006] Further, clustering each pixel according to the fault fitting degree and gray value of the fault edge where the pixel is located includes: Establish a pixel feature set according to the fault fitting degree and gray value of the fault edge where the pixel is located, and randomly select k initial clustering centers of the pixel feature set; Calculate the Euclidean distance from the pixel feature data in the pixel feature set to the initial clustering center, and divide each pixel into the corresponding clustering cluster according to the Euclidean distance from the pixel feature data in the pixel feature set to the initial clustering center; Calculate the mean value of the pixel feature data in each clustering cluster, and re-determine the clustering center according to the mean value of the pixel feature data in each clustering cluster; Repeat the above steps until the clustering center no longer changes or the number of iterations reaches the preset iteration threshold, and obtain the clustering results of each pixel.

[0007] Further, calculating the abnormality coefficient of the photovoltaic module segmentation image includes: Determine the average gray value and area value of each segmentation image according to the photovoltaic module segmentation image, and determine the abnormality coefficient of the photovoltaic module segmentation image based on the abnormality coefficient calculation formula. The abnormality coefficient calculation formula is specifically: , where, is the abnormality coefficient, is the average gray value of the photovoltaic module segmentation image, is the preset allowable gray value, is the preset conversion coefficient, is the area value of the photovoltaic module segmentation image, is the preset allowable area value, is the natural exponential function.

[0008] Further, screening out the abnormal images of the photovoltaic module according to the abnormality coefficient of the photovoltaic module segmentation image includes: Obtain the preset allowable abnormality coefficient, and calculate the difference between the current abnormality coefficient and the preset allowable abnormality coefficient; Judge whether the difference between the current abnormality coefficient and the preset allowable abnormality coefficient is greater than the first preset threshold. If the difference between the current abnormality coefficient and the preset allowable abnormality coefficient is greater than the first preset threshold, then determine the corresponding photovoltaic module segmentation image as an abnormal image; If the difference between the current abnormality coefficient and the preset allowable abnormality coefficient is less than or equal to the first preset threshold, then determine the corresponding photovoltaic module segmentation image as a normal image.

[0009] Further, determining the abnormality type of the photovoltaic module abnormal image according to the feature vector of the photovoltaic module abnormal image includes: Obtain historical photovoltaic module anomaly detection data, and determine the feature vectors and corresponding anomaly types of historical photovoltaic module anomaly images based on the historical photovoltaic module anomaly detection data; Preprocess the feature vectors and corresponding anomaly types of historical photovoltaic module anomaly images, establish a training sample set based on the preprocessed feature vectors and corresponding anomaly types of historical photovoltaic module anomaly images, and establish an initial anomaly evaluation model based on the training sample set; Train the initial anomaly evaluation model according to the training sample set to obtain a trained anomaly evaluation model; Input the feature vectors of the current photovoltaic module anomaly image into the trained anomaly evaluation model to obtain the corresponding anomaly type of the current photovoltaic module anomaly image.

[0010] Further, determining the warning weight according to the anomaly type of the photovoltaic module anomaly image includes: Obtain the preset standard area values corresponding to each anomaly type, and calculate the ratio of the area value of the current photovoltaic module anomaly image to the preset standard area value; Normalize the ratio of the area value of the current photovoltaic module anomaly image to the preset standard area value, and determine the warning weight according to the normalized ratio of the area value of the current photovoltaic module anomaly image to the preset standard area value.

[0011] Further, warning the photovoltaic module anomaly image according to the warning weight and the anomaly coefficient and locating the corresponding photovoltaic module includes: Multiply the warning weight by the anomaly coefficient of the corresponding photovoltaic module anomaly image to obtain the warning parameter of the photovoltaic module anomaly image.

[0012] Further, warning the photovoltaic module anomaly image according to the warning weight and the anomaly coefficient and locating the corresponding photovoltaic module further includes: Obtain the change situation of the warning parameter of the photovoltaic module anomaly image, and draw a warning parameter change curve according to the change situation of the warning parameter of the photovoltaic module anomaly image; Perform curve fitting on the warning parameter change curve to obtain a warning parameter prediction curve of the photovoltaic module anomaly image; Obtain the preset allowable warning parameter, and judge the time required for the warning parameter of the warning parameter prediction curve to reach the preset allowable warning parameter; If the time required for the warning parameter to reach the preset allowable warning parameter is less than the third preset threshold, then warn the photovoltaic module anomaly image and locate the corresponding photovoltaic module.

[0013] To achieve the above object, the present invention also provides a photovoltaic module anomaly location system, including: A segmentation module, configured to obtain infrared image data of a photovoltaic module, segment the infrared image data of the photovoltaic module, and obtain a plurality of segmented images of the photovoltaic module; A screening module, configured to calculate an anomaly coefficient of the segmented image of the photovoltaic module, and screen out an abnormal image of the photovoltaic module according to the anomaly coefficient of the segmented image of the photovoltaic module; A determination module, configured to obtain a feature vector of the abnormal image of the photovoltaic module, and determine the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; A positioning module, configured to determine an early warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and perform early warning on the abnormal image of the photovoltaic module according to the early warning weight and the anomaly coefficient, and locate the corresponding photovoltaic module.

[0014] The beneficial effects of the present invention are as follows: By applying the above technical solutions, the present invention obtains infrared image data of a photovoltaic module, segments the infrared image data of the photovoltaic module, and obtains a plurality of segmented images of the photovoltaic module; calculates the anomaly coefficient of the segmented image of the photovoltaic module, and screens out an abnormal image of the photovoltaic module according to the anomaly coefficient of the segmented image of the photovoltaic module; obtains a feature vector of the abnormal image of the photovoltaic module, and determines the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; determines an early warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and performs early warning on the abnormal image of the photovoltaic module according to the early warning weight and the anomaly coefficient, and locates the corresponding photovoltaic module. The present invention segments and warns and locates the abnormal photovoltaic module images through the anomaly coefficient and the early warning weight, can timely locate the abnormal photovoltaic area, and improves the maintenance efficiency of the photovoltaic module. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0016] Figure 1 Shows the overall flowchart of a method for abnormal positioning of a photovoltaic module proposed in an embodiment of the present invention; Figure 2 Shows the structural schematic diagram of a system for abnormal positioning of a photovoltaic module proposed in an embodiment of the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0018] An embodiment of the present application provides a method for abnormal positioning of a photovoltaic module, as Figure 1 shown, including: S101, obtaining infrared image data of a photovoltaic module, segmenting the infrared image data of the photovoltaic module to obtain a plurality of segmented images of the photovoltaic module; In some embodiments of the present application, the segmenting the infrared image data of the photovoltaic module to obtain a plurality of segmented images of the photovoltaic module includes: performing gray-scale processing on the infrared image data of the photovoltaic module to obtain a gray-scale image of the photovoltaic module; performing edge detection on the gray-scale image of the photovoltaic module based on the canny edge detection algorithm, and determining a fault edge according to the edge detection result of the gray-scale image of the photovoltaic module; obtaining a preset standard fault edge, and calculating a fault fitting degree between the fault edge of the current gray-scale image of the photovoltaic module and the preset standard fault edge; clustering each pixel point according to the fault fitting degree and gray value of the pixel point located at the fault edge, and connecting adjacent pixel points in the same clustering cluster according to the clustering result to obtain a segmented image of the photovoltaic module.

[0019] In this embodiment, by setting an infrared thermal imaging monitoring device near the power plant to monitor the infrared image data of the photovoltaic module in real time, the preset standard fault edge is specifically the fault edge of the gray-scale image when the photovoltaic module has a historical fault. The image is segmented by the fault fitting degree between the fault edge of the gray-scale image of the photovoltaic module and the preset standard fault edge and the pixel point gray value, which is convenient for subsequent segmentation of the abnormal image of the photovoltaic module.

[0020] In some embodiments of the present application, the clustering each pixel point according to the fault fitting degree and gray value of the pixel point located at the fault edge includes: establishing a pixel point feature set according to the fault fitting degree and gray value of the pixel point located at the fault edge, and randomly selecting k initial clustering centers of the pixel point feature set; calculating the Euclidean distance from the pixel point feature data in the pixel point feature set to the initial clustering center, and dividing each pixel point into the corresponding clustering cluster according to the Euclidean distance from the pixel point feature data in the pixel point feature set to the initial clustering center; calculating the mean value of the pixel point feature data in each clustering cluster, and re-determining the clustering center according to the mean value of the pixel point feature data in each clustering cluster; repeating the above steps iteratively until the clustering center no longer changes or the number of iterations reaches a preset iteration threshold to obtain the clustering result of each pixel point.

[0021] In this embodiment, pixel points are clustered based on the k-means clustering algorithm, and clustering clusters are assigned to pixel points according to the fault fitness and gray value.

[0022] S102. Calculate the anomaly coefficient of the segmented image of the photovoltaic module, and filter out the abnormal images of the photovoltaic module according to the anomaly coefficient of the segmented image of the photovoltaic module. In some embodiments of the present application, the calculating the anomaly coefficient of the segmented image of the photovoltaic module includes: determining the average gray value and area value of each segmented image according to the segmented image of the photovoltaic module, and determining the anomaly coefficient of the segmented image of the photovoltaic module based on the anomaly coefficient calculation formula. The anomaly coefficient calculation formula is specifically: , where, is the anomaly coefficient, is the average gray value of the segmented image of the photovoltaic module, is the preset allowable gray value, is the preset conversion coefficient, is the area value of the segmented image of the photovoltaic module, is the preset allowable area value, is the natural exponential function.

[0023] In some embodiments of the present application, the filtering out the abnormal images of the photovoltaic module according to the anomaly coefficient of the segmented image of the photovoltaic module includes: obtaining the preset allowable anomaly coefficient, and calculating the difference between the current anomaly coefficient and the preset allowable anomaly coefficient; determining whether the difference between the current anomaly coefficient and the preset allowable anomaly coefficient is greater than the first preset threshold. If the difference between the current anomaly coefficient and the preset allowable anomaly coefficient is greater than the first preset threshold, the corresponding segmented image of the photovoltaic module is determined as an abnormal image; if the difference between the current anomaly coefficient and the preset allowable anomaly coefficient is less than or equal to the first preset threshold, the corresponding segmented image of the photovoltaic module is determined as a normal image.

[0024] In this embodiment, the anomaly coefficient is calculated through the average gray value and area value of the segmented image of the photovoltaic module, and the abnormal images are filtered out by the difference between the corresponding anomaly coefficient of each segmented image of the photovoltaic module and the preset allowable anomaly coefficient, so as to realize the accurate extraction of the abnormal images.

[0025] S103. Obtain the feature vector of the abnormal image of the photovoltaic module, and determine the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module. In some embodiments of the present application, determining the abnormal type of the abnormal image of the photovoltaic module based on the feature vector of the abnormal image of the photovoltaic module includes: obtaining historical abnormal detection data of the photovoltaic module, and determining the feature vector and the corresponding abnormal type of the historical abnormal image of the photovoltaic module according to the historical abnormal detection data of the photovoltaic module; preprocessing the feature vector and the corresponding abnormal type of the historical abnormal image of the photovoltaic module, establishing a training sample set according to the preprocessed feature vector and the corresponding abnormal type of the historical abnormal image of the photovoltaic module, and establishing an initial abnormal evaluation model according to the training sample set; training the initial abnormal evaluation model according to the training sample set to obtain a trained abnormal evaluation model; inputting the feature vector of the current abnormal image of the photovoltaic module into the trained abnormal evaluation model to obtain the abnormal type corresponding to the current abnormal image of the photovoltaic module.

[0026] In this embodiment, an initial abnormal evaluation model is established based on a deep learning neural network model, and at the same time, the model is trained by establishing a training sample set with the feature vector and the corresponding abnormal type of the preprocessed historical abnormal image of the photovoltaic module, so as to output the abnormal type corresponding to the current abnormal image of the photovoltaic module through the trained abnormal evaluation model.

[0027] S104. Determine the warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and give a warning to the abnormal image of the photovoltaic module according to the warning weight and the abnormal coefficient, and locate the corresponding photovoltaic module.

[0028] In some embodiments of the present application, determining the warning weight according to the abnormal type of the abnormal image of the photovoltaic module includes: obtaining the preset standard area value corresponding to each abnormal type, and calculating the ratio of the area value of the current abnormal image of the photovoltaic module to the preset standard area value; normalizing the ratio of the area value of the current abnormal image of the photovoltaic module to the preset standard area value, and determining the warning weight according to the normalized ratio of the area value of the current abnormal image of the photovoltaic module to the preset standard area value.

[0029] In this embodiment, the system has a preset standard area value corresponding to each abnormal type, and the warning weight is obtained through the ratio of the area value of the current abnormal image of the photovoltaic module to the preset standard area value.

[0030] In some embodiments of the present application, warning the abnormal image of the photovoltaic module according to the warning weight and the abnormal coefficient and locating the corresponding photovoltaic module includes: multiplying the warning weight by the abnormal coefficient of the corresponding abnormal image of the photovoltaic module to obtain the warning parameter of the abnormal image of the photovoltaic module.

[0031] In some embodiments of the present application, the method of warning about the abnormal image of a photovoltaic module according to the warning weight and the abnormal coefficient and locating the corresponding photovoltaic module further includes: obtaining the change situation of the warning parameters of the abnormal image of the photovoltaic module, and drawing a warning parameter change curve according to the change situation of the warning parameters of the abnormal image of the photovoltaic module; performing curve fitting on the warning parameter change curve to obtain a warning parameter prediction curve of the abnormal image of the photovoltaic module; obtaining a preset allowable warning parameter, and judging the time required for the warning parameter of the warning parameter prediction curve to reach the preset allowable warning parameter; if the time required for the warning parameter to reach the preset allowable warning parameter is less than a third preset threshold, warning about the abnormal image of the photovoltaic module and locating the corresponding photovoltaic module.

[0032] In this embodiment, curve fitting is performed on the warning parameter change curve based on the least squares method to obtain a warning parameter prediction curve, so as to give a warning in time and locate the corresponding photovoltaic module when the warning parameter of the warning parameter prediction curve is about to reach the preset allowable warning parameter.

[0033] Based on the same technical concept, as Figure 2 shown, the present invention also provides a photovoltaic module abnormal location system, including: A segmentation module, configured to obtain infrared image data of a photovoltaic module, segment the infrared image data of the photovoltaic module to obtain a plurality of segmented photovoltaic module images; a screening module, configured to calculate the abnormal coefficient of the segmented photovoltaic module images, and screen out the abnormal images of the photovoltaic module according to the abnormal coefficient of the segmented photovoltaic module images; a judgment module, configured to obtain the feature vector of the abnormal image of the photovoltaic module, and determine the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; a location module, configured to determine the warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and warn about the abnormal image of the photovoltaic module according to the warning weight and the abnormal coefficient and locate the corresponding photovoltaic module.

[0034] By applying the above technical solution, the present invention obtains the infrared image data of the photovoltaic module, segments the infrared image data of the photovoltaic module to obtain a plurality of segmented photovoltaic module images; calculates the abnormal coefficient of the segmented photovoltaic module images, and screens out the abnormal images of the photovoltaic module according to the abnormal coefficient of the segmented photovoltaic module images; obtains the feature vector of the abnormal image of the photovoltaic module, and determines the abnormal type of the abnormal image of the photovoltaic module according to the feature vector of the abnormal image of the photovoltaic module; determines the warning weight according to the abnormal type of the abnormal image of the photovoltaic module, and warns about the abnormal image of the photovoltaic module according to the warning weight and the abnormal coefficient and locates the corresponding photovoltaic module. The present invention segments the abnormal photovoltaic module images through the abnormal coefficient and the warning weight and performs warning and location, and can timely locate the abnormal photovoltaic area, improving the maintenance efficiency of the photovoltaic module.

[0035] Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by hardware or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various implementation scenarios of the present invention.

[0036] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for locating abnormal photovoltaic components, characterized in that: The method comprises: Acquire infrared image data of photovoltaic modules, segment the infrared image data of photovoltaic modules, and obtain a plurality of segmented images of photovoltaic modules; Calculate the abnormal coefficient of the photovoltaic component segmentation image, and screen out abnormal photovoltaic component images according to the abnormal coefficient of the photovoltaic component segmentation image; Acquire a feature vector of an abnormal image of a photovoltaic component, and determine an abnormal type of the abnormal image of the photovoltaic component according to the feature vector of the abnormal image of the photovoltaic component; The warning weight is determined according to the abnormal type of the abnormal image of the photovoltaic component, and the abnormal image of the photovoltaic component is warned according to the warning weight and the abnormal coefficient, and the corresponding photovoltaic component is located.

2. The photovoltaic module abnormality positioning method according to claim 1, characterized in that: The photovoltaic module infrared image data is segmented to obtain a plurality of photovoltaic module segmented images, including: Grayscale processing is performed on the infrared image data of the photovoltaic module to obtain a grayscale image of the photovoltaic module; Perform edge detection on the grayscale image of the photovoltaic module based on the canny edge detection algorithm, and determine the fault edge according to the edge detection result of the grayscale image of the photovoltaic module; Obtaining a preset standard fault edge, and calculating the fault fit between the fault edge of the current photovoltaic module grayscale image and the preset standard fault edge; Each pixel is clustered according to the fault fitting degree and grayscale value of the fault edge where the pixel is located. According to the clustering result, the adjacent pixels in the same cluster are connected to obtain the photovoltaic module segmentation image.

3. The photovoltaic module abnormality positioning method according to claim 2, characterized in that: The clustering of each pixel point according to the fault fitting degree and gray value of the fault edge where the pixel point is located includes: A pixel feature set is established according to the fault fitting degree and gray value of the fault edge where the pixel is located, and k initial clustering centers of the pixel feature set are randomly selected; Calculate the Euclidean distance from the pixel feature data in the pixel feature set to the initial cluster center, and divide each pixel into a corresponding cluster according to the Euclidean distance from the pixel feature data in the pixel feature set to the initial cluster center; Calculate the mean value of the feature data of the pixels in each cluster, and re-determine the cluster center according to the mean value of the feature data of the pixels in each cluster; Repeat the above steps until the cluster center no longer changes or the number of iterations reaches a preset iteration threshold, and obtain the clustering results of each pixel.

4. The photovoltaic module abnormality positioning method according to claim 1, characterized in that: The step of calculating the abnormal coefficient of the photovoltaic component segmentation image comprises: The average grayscale value and area value of each segmented image are determined according to the segmented image of the photovoltaic component, and the abnormal coefficient of the segmented image of the photovoltaic component is determined according to the average grayscale value and area value of the segmented image based on the abnormal coefficient calculation formula. The abnormal coefficient calculation formula is specifically: ， in, is the abnormal coefficient, is the average gray value of the PV module segmentation image, is the preset allowable gray value, is the preset conversion factor, is the area value of the PV module segmentation image, is the preset allowable area value, is a natural exponential function.

5. The photovoltaic module abnormality positioning method according to claim 4, characterized in that: The step of screening out abnormal images of photovoltaic components according to abnormal coefficients of the photovoltaic component segmented images includes: Obtaining a preset allowable abnormality coefficient, and calculating the difference between the current abnormality coefficient and the preset allowable abnormality coefficient; Determine whether the difference between the current abnormal coefficient and the preset allowable abnormal coefficient is greater than a first preset threshold value, and if the difference between the current abnormal coefficient and the preset allowable abnormal coefficient is greater than the first preset threshold value, determine the corresponding photovoltaic component segmented image as an abnormal image; If the difference between the current abnormal coefficient and the preset allowable abnormal coefficient is less than or equal to the first preset threshold, the corresponding photovoltaic component segmented image is determined as a normal image.

6. The photovoltaic module abnormality positioning method according to claim 1, characterized in that: The determining the abnormal type of the abnormal image of the photovoltaic component according to the feature vector of the abnormal image of the photovoltaic component includes: Acquire historical photovoltaic component anomaly detection data, and determine the feature vector of the historical photovoltaic component abnormal image and the corresponding abnormal type according to the historical photovoltaic component anomaly detection data; Preprocessing the feature vectors of historical photovoltaic component abnormal images and the corresponding abnormal types, establishing a training sample set based on the preprocessed feature vectors of historical photovoltaic component abnormal images and the corresponding abnormal types, and establishing an initial abnormality assessment model based on the training sample set; The initial anomaly assessment model is trained according to the training sample set to obtain a trained anomaly assessment model; The feature vector of the current abnormal image of the photovoltaic component is input into the trained abnormal assessment model to obtain the abnormal type corresponding to the current abnormal image of the photovoltaic component.

7. The photovoltaic module abnormality positioning method according to claim 6, characterized in that: The step of determining the warning weight according to the abnormal type of the abnormal image of the photovoltaic component includes: Obtaining the preset standard area value corresponding to each abnormal type, and calculating the ratio of the area value of the current abnormal image of the photovoltaic component to the preset standard area value; The ratio of the area value of the current abnormal image of the photovoltaic component to the preset standard area value is normalized, and the warning weight is determined according to the ratio of the area value of the current abnormal image of the photovoltaic component after the normalization to the preset standard area value.

8. The photovoltaic module abnormality positioning method according to claim 1, characterized in that: The method of issuing an early warning for abnormal images of photovoltaic components and locating corresponding photovoltaic components according to the early warning weight and the abnormal coefficient includes: The warning weight is multiplied by the abnormal coefficient of the corresponding PV module abnormal image to obtain the warning parameter of the PV module abnormal image.

9. The photovoltaic module abnormality positioning method according to claim 8, characterized in that: The method of issuing an early warning for abnormal images of photovoltaic components and locating corresponding photovoltaic components according to the early warning weight and the abnormal coefficient also includes: Obtain the change of warning parameters of abnormal images of photovoltaic components, and draw a warning parameter change curve according to the change of warning parameters of abnormal images of photovoltaic components; Perform curve fitting on the early warning parameter change curve to obtain the early warning parameter prediction curve of the abnormal image of the photovoltaic module; Obtaining preset allowable warning parameters, and determining the time required for the warning parameters of the warning parameter prediction curve to reach the preset allowable warning parameters; If the time required for the warning parameter to reach the preset allowable warning parameter is less than the third preset threshold, an abnormal image of the photovoltaic component is warned and the corresponding photovoltaic component is located.

10. A photovoltaic module abnormality positioning system, characterized in that: include: A segmentation module is used to obtain infrared image data of photovoltaic modules, segment the infrared image data of photovoltaic modules, and obtain a plurality of segmented images of photovoltaic modules; A screening module is used to calculate an abnormal coefficient of the photovoltaic component segmentation image, and screen out abnormal photovoltaic component images according to the abnormal coefficient of the photovoltaic component segmentation image; A judgment module is used to obtain a feature vector of an abnormal image of a photovoltaic component, and determine an abnormal type of the abnormal image of the photovoltaic component according to the feature vector of the abnormal image of the photovoltaic component; The positioning module is used to determine the warning weight according to the abnormal type of the abnormal image of the photovoltaic component, and to issue a warning to the abnormal image of the photovoltaic component according to the warning weight and the abnormal coefficient and to locate the corresponding photovoltaic component.

Images