Hot spot fault detection method and device for photovoltaic module, equipment and storage medium

Abstract

The invention discloses a hot spot fault detection method, device and equipment for a photovoltaic module and a readable storage medium, and the method comprises the steps: carrying out the dual-light image collection of a photovoltaic power station through an unmanned plane, obtaining an infrared image and a visible light image, and generating an infrared panoramic image and a visible light panoramic image respectively; according to the temperature parameters corresponding to all the pixel points in the infrared panoramic image, iron oxide red color mode rendering and thermal color mode rendering are carried out on the infrared panoramic image, and an iron oxide red color image and a thermal color image are generated; calibrating an imaging position of the hot spot battery and a hot spot fault type according to the iron oxide red color image and the heating power color image; and displaying and outputting the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type according to the geographic position information of the imaging position of the hot spot battery in the infrared panoramic image. Simple and rapid operation and maintenance of the photovoltaic power station are realized, and the operation and maintenance cost of the power station is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station operation and maintenance, and particularly to a method, device, equipment and computer-readable storage medium for detecting hot spot faults of photovoltaic modules. Background Art

[0002] With the rapid development of the photovoltaic industry, the area scale of photovoltaic power stations is increasing day by day; and the operation and maintenance of photovoltaic power stations is crucial for the high-efficiency operation of photovoltaic power stations, which can effectively improve the economic benefits of photovoltaic power stations and avoid the occurrence of safety accidents in photovoltaic power stations. However, large-scale photovoltaic power stations pose challenges to the operation and maintenance of power stations, which require huge manual time costs and are difficult to maintain.

[0003] With the development of unmanned aerial vehicle (UAV) technology, using UAVs equipped with infrared thermal imagers or cameras to scan and collect images of photovoltaic modules in photovoltaic power stations at low altitudes, and using image analysis and recognition technology to intelligently identify faulty photovoltaic cells for photovoltaic applications in photovoltaic power station inspections, which greatly reduces the labor costs consumed in photovoltaic power station inspections. However, to accurately identify hot spot cells in photovoltaic module images, relatively complex image processing technologies are often required, with high computational difficulty and higher equipment costs for computational processing, which also increases the operation and maintenance costs of photovoltaic power stations. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, device, equipment and computer-readable storage medium for detecting hot spot faults of photovoltaic modules, which can reduce the operation and maintenance costs on the basis of realizing simple and rapid operation and maintenance of photovoltaic power stations.

[0005] To solve the above technical problems, the present invention provides a method for detecting hot spot faults of photovoltaic modules, including:

[0006] Performing dual-light image acquisition on a photovoltaic power station by using a UAV to obtain an infrared image and a visible light image, and respectively generating an infrared panoramic image and a visible light panoramic image;

[0007] Rendering the infrared panoramic image in an iron-red color mode and a thermal color mode respectively according to the temperature parameters corresponding to each pixel point in the infrared panoramic image to generate an iron-red color image and a thermal color image;

[0008] Calibrating the imaging position and the type of hot spot fault of the hot spot cell according to the iron-red color image and the thermal color image;

[0009] Displaying and outputting the distribution position and the type of hot spot fault of the hot spot cell in the visible light panoramic image according to the geographical location information of the imaging position of the hot spot cell in the infrared panoramic image.

[0010] In an alternative embodiment of the present application, the process of generating the infrared panoramic image includes:

[0011] Generating a raw file containing temperature information data based on the infrared image;

[0012] Performing numerical calculation conversion on the temperature information data in the raw file to obtain the temperature parameters corresponding to the infrared image;

[0013] Extracting the geographical location information corresponding to the infrared image according to the position and attitude information of the drone when taking each infrared image;

[0014] Converting each infrared image into a TIFF format image with the temperature parameters and geographical location information according to the geographical location information;

[0015] Performing image stitching on each TIFF format image to obtain an infrared panoramic image.

[0016] In an alternative embodiment of the present application, the process of rendering the infrared panoramic image according to the temperature information data of the infrared panoramic image to generate an iron red color image and a thermal color image includes:

[0017] Forming a temperature parameter set with the temperature parameters of each pixel point in the infrared panoramic image as set elements;

[0018] Dividing each temperature parameter element in the temperature parameter set into multiple temperature grading sets;

[0019] Using the natural breakpoint method formula , combined with each temperature parameter element in the temperature parameter set, for and Performing iterative optimization adjustment to obtain multiple temperature grading sets obtained by dividing each temperature parameter element when the variance goodness of fit is the largest and not less than the set goodness of fit threshold; where is the variance goodness of fit, is the th temperature parameter element in the temperature parameter set; is the average value of each temperature parameter element in the temperature parameter set; is the number of the temperature grading sets; is the th is the th th temperature parameter element in the th temperature grading set, the average value of each of the temperature parameter elements in each of the temperature classification sets; and , ;

[0020] Taking the maximum temperature parameter element and the minimum temperature parameter in each of the temperature classification sets as the primary color temperature parameter interval;

[0021] Adopting the iron red color mode, for the infrared panoramic image, pixels corresponding to the temperature parameter elements belonging to the same level of the color temperature parameter interval are assigned the same color parameter, and pixels corresponding to the temperature parameter elements belonging to different levels of the color temperature parameter interval are assigned different color parameters for rendering to obtain the iron red color rendered image;

[0022] Adopting the thermal color mode, for the infrared panoramic image, pixels corresponding to the temperature parameter elements belonging to the same level of the color temperature parameter interval are assigned the same color parameter, and pixels corresponding to the temperature parameter elements belonging to different levels of the color temperature parameter interval are assigned different color parameters for rendering to obtain the thermal color rendered image.

[0023] In an optional embodiment of the present application, the process of generating the visible light panoramic image includes:

[0024] Stitching the visible light images through Context Capture software to build the formed visible light panoramic image; wherein, the height at which the drone collects the visible light images and the infrared images is not higher than 40m, and the overlap rate between two adjacent frames of the visible light images along the heading or side direction of the drone is not less than 70%.

[0025] In an optional embodiment of the present application, after separately generating the visible light panoramic image and the infrared panoramic image, it further includes:

[0026] Importing the visible light panoramic image and the infrared panoramic image into the LocaSpace Viewer software together, and determining the corresponding relationship between the pixel points in the visible light panoramic image and the pixel points in the infrared panoramic image according to the geographical location information respectively corresponding to the visible light panoramic image and the infrared panoramic image;

[0027] Correspondingly, according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image, displaying and outputting the distribution position and the hot spot fault type of the hot spot battery in the visible light panoramic image, including:

[0028] Based on the geographical location information of the hot-spot cell in the infrared panoramic image and the corresponding relationship, display and output the distribution position and the hot-spot fault type of the hot-spot cell in the visible-light panoramic image.

[0029] In an optional embodiment of the present application, after calibrating the imaging position and the hot-spot fault type of the hot-spot cell according to the iron-red color image and the thermal-color image, it further includes:

[0030] According to the number of hot-spot cell strings in each of the hot-spot defective components, divide the first fault category of the hot-spot defective components into single-string cell faults, two-string cell faults, and three-string cell faults;

[0031] According to the distribution positions of the hot-spot cells in each of the hot-spot cell strings, divide the second fault category of each of the hot-spot cell strings into dot faults, dot-strip faults, and strip faults;

[0032] Correspondingly, displaying and outputting the distribution position and the hot-spot fault type of the hot-spot cell in the visible-light panoramic image includes:

[0033] Output the corresponding first fault category, second fault type, distribution position, and hot-spot fault type of each of the hot-spot defective components; wherein, the hot-spot fault type includes occluded hot-spot faults and non-occluded hot-spot faults.

[0034] A fault detection device for a photovoltaic module, comprising:

[0035] An image processing module, configured to generate an infrared panoramic image and a visible-light panoramic image respectively according to infrared images and visible-light images obtained by a drone for dual-light image acquisition of a photovoltaic power station;

[0036] A rendering display module, configured to perform iron-red color mode rendering and thermal-color mode rendering on the infrared panoramic image respectively according to the temperature information data corresponding to each pixel point in the infrared panoramic image, and generate an iron-red color image and a thermal-color image;

[0037] A fault calibration module, configured to calibrate the imaging position and the hot-spot fault type of the hot-spot cell according to the iron-red color image and the thermal-color image;

[0038] An information output module, configured to display and output the distribution position and the hot-spot fault type of the hot-spot cell in the visible-light panoramic image according to the geographical location information of the imaging position of the hot-spot cell in the infrared panoramic image.

[0039] In an optional embodiment of the present application, the image processing module includes an infrared processing unit, which is configured to generate a raw file containing temperature information data according to the infrared image; perform numerical calculation conversion on the temperature information data in the raw file to obtain the temperature parameters corresponding to the infrared image; extract the geographical location information corresponding to the infrared image according to the position and attitude information of the drone when taking each infrared image; convert each infrared image into a TIFF format image with the temperature parameters and geographical location information according to the geographical location information; and splice the TIFF format images to obtain an infrared panoramic image.

[0040] A fault detection device for a photovoltaic module, comprising:

[0041] A memory for storing a computer program;

[0042] A processor for executing the computer program to implement the steps of the fault detection method for a photovoltaic module as described in any one of the above.

[0043] A computer-readable storage medium stores a computer program, and the computer program is executed to implement the steps of the fault detection method for a photovoltaic module as described in any one of the above.

[0044] A hot spot fault detection method, device, equipment and readable storage medium for a photovoltaic module provided by the present invention. The hot spot fault detection method for the photovoltaic module includes obtaining an infrared image and a visible light image by performing dual-light image acquisition on a photovoltaic power station by a drone, and respectively generating an infrared panoramic image and a visible light panoramic image; performing iron-red color mode rendering and thermal color mode rendering on the infrared panoramic image respectively according to the temperature parameters corresponding to each pixel point in the infrared panoramic image to generate an iron-red color image and a thermal color image; calibrating the imaging position and the hot spot fault type of the hot spot battery according to the iron-red color image and the thermal color image; and displaying and outputting the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image.

[0045] In this application, an infrared panoramic image and a visible light panoramic image of a photovoltaic power station are obtained by simultaneous scanning using a drone; on this basis, the infrared panoramic image with temperature parameters is rendered and displayed in two different rendering modes; thus, users can directly and more simply and quickly identify and calibrate hot spot cells with hot spot phenomena caused by different reasons based on the images rendered in the two different rendering modes, so as to accurately and comprehensively identify the existing hot spot cells, and after determining the hot spot cells, further locate the position where the hot spot cells are located with the help of the visible light panoramic image, thereby reducing the difficulty of maintaining and overhauling the hot spot cells. It can be seen that this application realizes simple and rapid operation and maintenance of the photovoltaic power station and reduces the operation and maintenance cost of the power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is a schematic flow chart of a method for detecting hot spot faults of a photovoltaic module provided by an embodiment of the present application;

[0048] Figure 2 It is a schematic diagram of the classification of the fault categories of hot spot defect modules provided by an embodiment of the present application;

[0049] Figure 3 It is a schematic diagram of the fault category marking of hot spot defect modules provided by an embodiment of the present application;

[0050] Figure 4 It is a structural block diagram of a device for detecting hot spot faults of a photovoltaic module provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The core of the present invention is to provide a method, device, equipment and readable storage medium for detecting hot spot faults of a photovoltaic module, which reduces the difficulty of identifying and calculating hot spot faults of the photovoltaic module to a certain extent and improves the convenience of operation and maintenance of the photovoltaic module.

[0052] In order to enable those skilled in the art to better understand the solution of the present invention, the following will further elaborate on the present invention in conjunction with the drawings and specific embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0053] Such asFigure 1 As shown in Figure 1 This is a schematic flowchart of a method for detecting hot spot faults of a photovoltaic module provided by an embodiment of the present application.

[0054] In a specific embodiment of the present application, the method for detecting hot spot faults of the photovoltaic module may include:

[0055] S1: Obtain infrared images and visible light images by performing dual - light image acquisition on a photovoltaic power station using a drone, and generate an infrared panoramic image and a visible light panoramic image respectively.

[0056] In this embodiment, during the process of using the drone for dual - light image acquisition, a drone equipped with both an infrared camera and a visible light camera can fly over the photovoltaic power station, so as to synchronously acquire the infrared images and visible light images of the photovoltaic power station. Based on the infrared images acquired during the flight of the drone, an infrared panoramic image can be formed by stitching; and each visible light image can be stitched to form a visible light panoramic image.

[0057] In addition, the flight height of the drone will directly affect the image stitching. If the flight height of the drone is too high, the clarity of the infrared image will decrease, resulting in blurred images without clear textures, which will affect the image stitching effect and success rate. Moreover, the too - high shooting distance of the drone will also lead to inaccurate geographical location information of the photos, thus affecting the subsequent positioning of the faulty photovoltaic modules. However, if the flight height of the drone is too low and the overlap rate is too high, the number of photos taken will increase significantly, and the stitching time will also increase substantially. To ensure the usability of the stitching effect and reduce the flight time and stitching time, when the drone acquires visible light images and infrared images, the flight height is not higher than 40m, and the overlap rate between two adjacent visible light images to be stitched along the heading or side - looking direction of the drone is not less than 70%. Currently, in the industry, drones with two resolutions of 640*512 and 1280*1024 are mainly used to carry lenses for power station inspection. When using a drone with a resolution of 640*512 for inspection operations, the flight height is set not to exceed 30m, and the overlap rates of the acquired images in the heading and side - looking directions are set not less than 80% and 70% respectively; when using a drone with a resolution of 1280*1024 for inspection operations, the flight height is set not to exceed 40m, and the overlap rates of the acquired images in the heading and side - looking directions are set not less than 80% and 70% respectively, and finally, dual - light images that can be stitched are output.

[0058] On this basis, during the image stitching process for the dual - light images taken by the drone, the visible light images can be used to build a model through Context Capture software to export a 2D visible light panoramic base map.

[0059] And the process of generating an infrared panoramic image based on the infrared images may include:

[0060] S11: Generate a raw file containing temperature information data based on the infrared image;

[0061] S12: Numerically calculate and convert the temperature information data in the raw file to obtain the temperature parameters corresponding to the infrared image;

[0062] S13: Extract the geographical location information corresponding to the infrared image according to the position and attitude information when the drone captures each infrared image;

[0063] S14: Convert each infrared image into a TIFF format image with temperature parameters and geographical location information according to the geographical location information;

[0064] S15: Stitch the TIFF format images to obtain an infrared panoramic image.

[0065] In practical applications, the infrared images captured by the drone can be pre - processed first. Convert the infrared images in jpg format to TIFF format. In this process, a raw file storing temperature data can be generated based on the original infrared image by calling the dji - tsdk script. Then, numerically calculate and convert the temperature data in the raw file to obtain the temperature value. Next, extract the geographical location information corresponding to the original infrared image, rewrite the temperature value, geographical location information, etc., and then convert the format of the original infrared image to TIFF format. Finally, output a TIFF format image with information data such as the temperature information data and geographical location information of the original infrared image. Import the TIFF format image data into the pix4d software for panoramic base map stitching. During the stitching process, geographical location information such as longitude, latitude, altitude, and flight deviation angle should be retained.

[0066] S2: Render the infrared panoramic image in iron - red color mode and thermal - color mode respectively according to the temperature parameters corresponding to each pixel point in the infrared panoramic image to generate an iron - red color image and a thermal - color image.

[0067] The temperature parameters in this embodiment can specifically be the gray values of each pixel point in the infrared panoramic image, or the temperature values determined by converting based on the gray values. In short, as long as they can represent the temperature corresponding to each pixel point.

[0068] The gray value of each pixel point in the initial infrared panoramic image is also determined based on the temperature value corresponding to that pixel point, making the temperature color area in the color base map of the infrared panoramic image appear chaotic and unable to highlight the hot - spot defects of the photovoltaic modules; it is also difficult for users to clearly and separately identify the hot - spot cells with hot - spot faults from the infrared panoramic image.

[0069] To this end, in this embodiment, the stitched infrared panoramic image is color-adjusted and optimized, so that the imaging area of ​​the hot-spot battery is more prominent in the infrared panoramic image, so that the user can identify the hot-spot battery more easily and clearly.

[0070] In order to better highlight the imaging area of ​​the hot spot battery in this embodiment, two different color modes, iron red color mode and thermal color mode, are used to render and display the infrared panoramic image. In an optional implementation of this embodiment, the process of generating the iron red color image and the thermal color image may include:

[0071] S21: Taking the temperature parameter of each pixel in the infrared panoramic image as a set element, a temperature parameter set is formed.

[0072] The temperature parameter in this embodiment may be a grayscale value or a temperature value corresponding to each pixel in the infrared panoramic image, and each temperature parameter element in the temperature parameter set may be arranged from small to large or from large to small.

[0073] S22: Divide each temperature parameter element in the temperature parameter set into a plurality of temperature classification sets.

[0074] This step is equivalent to dividing several adjacent temperature parameter elements in the temperature parameter set into the same temperature classification set. Each temperature classification set should contain at least one temperature parameter element, and the number of temperature parameter elements in each temperature classification set is not necessarily the same.

[0075] S23: Using the natural breakpoint method formula , combined with each temperature parameter element in the temperature parameter set, and Iterative optimization and adjustment are performed to obtain multiple temperature classification sets for each temperature parameter element when the variance goodness of fit is maximum and not less than the set goodness of fit threshold.

[0076] in, is the variance goodness of fit, is the first temperature parameter in the set Temperature parameter elements; is the average value of each temperature parameter element in the temperature parameter set; The number of temperature classification sets; For the The number of temperature parameter elements in a temperature classification set; For the The first temperature classification set Temperature parameter elements, For the The average value of each temperature parameter element in the temperature classification set; and , .

[0077] S24: Use the maximum temperature parameter element and the minimum temperature parameter in each temperature grading set as the first-level color temperature parameter interval.

[0078] S25: Adopt the iron red color mode. For the infrared panoramic image, assign the same color parameter to the pixel points whose corresponding temperature parameter elements belong to the same first-level color temperature parameter interval, and assign different color parameters to the pixel points whose corresponding temperature parameter elements belong to different first-level color temperature parameter intervals for rendering to obtain an iron red color-rendered image;

[0079] S26: Adopt the thermal color mode. For the infrared panoramic image, assign the same color parameter to the pixel points whose corresponding temperature parameter elements belong to the same first-level color temperature parameter interval, and assign different color parameters to the pixel points whose corresponding temperature parameter elements belong to different first-level color temperature parameter intervals for rendering to obtain a thermal color-rendered image.

[0080] For the infrared panoramic image with chaotic temperature color area performance in the color map, optimize the color parameters by the natural breaks method. The natural breaks method is a statistical method for classification and grading according to the numerical statistical distribution law. By finding the natural turning points and characteristic points in the temperature information data as the grading boundaries for rendering colors, the temperature colors can be divided into different levels through the natural breaks algorithm, minimizing the temperature differences within each temperature color level, thus highlighting the maximum temperature differences between each temperature color level, and effectively highlighting the color performance of each temperature color level, facilitating the user to perform visual recognition and analysis on the infrared panoramic image.

[0081] In the process of re-dividing the temperature color levels by the natural breaks algorithm, the temperature parameter of each pixel point referred to in this embodiment is determined based on the temperature information data of each pixel point originally carried in the infrared panoramic image. Specifically, it can directly be the temperature value corresponding to each pixel point, or it can be the gray value corresponding to each pixel point. In short, it is the parameter data that can represent the temperature information corresponding to each pixel point.

[0082] In the infrared panoramic image, there must be at least two pixel points with the same corresponding temperature parameter. Based on the characteristic that each element in the set is different, in the temperature parameter set formed by the temperature parameters of all pixel points in the infrared panoramic image, each temperature parameter element must be different. This temperature parameter set can be expressed as , which is the total number of elements in the temperature parameter set; the average value of each temperature parameter element in the temperature parameter set is ; the total sum of squared deviations of all temperature parameter elements in this temperature parameter set is .

[0083] Divide each temperature parameter element in the temperature parameter set into groups, wherein adjacent temperature parameter elements are divided into a group, and each group of temperature parameter elements constitutes a temperature classification set; and The number of temperature parameter elements in a temperature classification set is , where the number of temperature parameter elements in each temperature classification set is not necessarily the same; Taking the example that the temperature parameter elements in the temperature parameter set are arranged in ascending order according to the corresponding temperature values, the temperature parameter elements in each temperature classification set are also arranged in ascending order accordingly.

[0084] Therefore, the first temperature classification set can be expressed as ; That is, the first to the second temperature parameter set A set of temperature parameter elements. The temperature classification set can be expressed as ;No. The temperature classification set can be expressed as .

[0085] Because the temperature parameter elements in each temperature classification set are also arranged from small to large according to the corresponding temperature values; therefore, .

[0086] In addition, the average value of the temperature parameter elements in each temperature grouping set can be expressed as: ; The cumulative value of the sum of squares of the deviations of the temperature parameter elements corresponding to each temperature grouping set is ;

[0087] In mathematics, the goodness of fit of variance is often used to test the rationality of classification. The goodness of fit of variance is also .

[0088] The number of temperature classification sets formed by dividing each temperature parameter element in the temperature parameter set through repeated optimization and iteration , and the number of temperature parameter elements in each temperature classification set ; Finally, a set of and Make the variance fit goodness of Closest to 1 and greater than the set goodness-of-fit threshold (avoiding variance goodness-of-fit too small), at this time, the first temperature parameter element and the last temperature parameter element in each temperature classification set correspond to the minimum temperature parameter element and the maximum temperature parameter element in each primary color temperature parameter interval, that is, the temperature parameter boundary value corresponding to a temperature color level.

[0089] After determining the color temperature parameter interval corresponding to each temperature color level, the infrared panoramic image can be further rendered and displayed in two different modes based on the color temperature parameter intervals at each level.

[0090] The iron-red color rendering image generated by rendering the infrared panoramic image in iron-red color mode can better highlight the location of the hot spot cells. The higher the color saturation and the brighter the color, the higher the temperature; and the lower the color brightness, the lower the temperature. This helps users find hot spots in the iron-red color rendering image that are obviously brighter than other areas in the imaging area of ​​a single photovoltaic module or the imaging areas of other surrounding photovoltaic modules. These are identified as hot spot defective components. This type of hot spot defective components are mostly hot spots caused by failures in the photovoltaic cells or circuit structures in the components, that is, non-obstructed failures.

[0091] The thermal color rendering image generated by rendering the infrared panoramic image using the thermal color mode is more conducive to highlighting the hot spots on the surface of the photovoltaic module caused by large-scale occlusion such as dust and vegetation. When performing thermal color rendering on the infrared panoramic image, the rendering color of each pixel can be arranged in the order of seven colors. The redder the color, the higher the temperature, followed by orange, and purple indicates the lowest temperature. When there are areas in the thermal color rendering image that are obviously more orange or red, it can be preferentially judged that the surface of the photovoltaic module in the area is blocked by dust, vegetation, etc., resulting in a higher temperature than the photovoltaic modules in other areas; this special highlighting form can intuitively display the distribution of temperature color data in the thermal color rendering image in the region and geographical area.

[0092] S3: Calibrate the imaging position of the hot spot battery and the hot spot fault type according to the iron red color image and the thermal color image.

[0093] After generating infrared panoramic images rendered in two different color modes, the two infrared panoramic images in different rendering modes can be displayed separately on a display. The user can simply and clearly determine the imaging position of the hot spot battery where the hot spot exists only according to the color rendered in the infrared panoramic image displayed on the display, and select the hot spot battery by clicking or circling, thereby realizing the calibration of the hot spot battery. This method of identifying the hot spot battery does not require complex image recognition calculation process, and can greatly reduce the difficulty of identifying hot spot faulty components.

[0094] Of course, in practical applications, image processing technology can also be used to identify hot-spot cells in the iron-red color image and the thermal color image respectively. Of course, considering the large amount of data in the infrared panoramic image, when actually identifying hot-spot cells in the infrared panoramic image, the hot-spot cells can be identified for each small imaging area in the infrared panoramic image in turn. In short, as long as the hot-spot cells can be identified with the help of two different infrared panoramic images, namely the iron-red color image and the thermal color image.

[0095] In addition, after generating infrared panoramic images rendered in two different color modes respectively, the visible light panoramic image and the infrared panoramic image can be further fused. Specifically, the visible light panoramic image and the infrared panoramic image can be jointly imported into the LocaSpace Viewer software, and the corresponding relationship between the pixel points in the visible light panoramic image and the pixel points in the infrared panoramic image can be determined according to the geographical location information corresponding to the visible light panoramic image and the infrared panoramic image respectively.

[0096] Since the visible light panoramic image and the infrared panoramic image are images generated by scanning the same photovoltaic power station scene by a drone, the geographical location information of each photovoltaic module imaging area in the visible light panoramic image and the geographical location information of the imaging positions of each photovoltaic module in the infrared panoramic image should correspond one by one.

[0097] In practical applications, switching displays can be performed between the visible light panoramic image and the infrared panoramic image, and also between the infrared panoramic images rendered in two different color modes. When it is necessary to calibrate the hot-spot cells with hot-spot phenomena through the infrared panoramic image, the iron-red color image in the infrared panoramic image can be first displayed on the display interface of the monitor. After calibrating the hot-spot cells based on the iron-red color image, the image displayed on the display interface can be switched to the thermal color image, and the hot-spot cells can be calibrated again based on this thermal color image. Because there are photovoltaic cells with the same geographical location information in the visible light panoramic image for each photovoltaic cell in the infrared panoramic image; when a hot-spot cell is selected in the infrared panoramic image, the currently displayed infrared panoramic image can be switched to the visible light panoramic image, and according to the geographical location information corresponding to the hot-spot cell and the corresponding relationship between the pixel points in the visible light panoramic image and the pixel points in the infrared panoramic image, the imaging position of the hot-spot cell in the visible light panoramic image can be displayed.

[0098] Different from the fact that the imaging area corresponding to the hot-spot cells can be highlighted in the infrared panoramic image, the visible light panoramic image can more realistically display the actual landform scene of the photovoltaic power station, and thus can more accurately display the distribution position of the hot-spot cells in the photovoltaic power station for users, thereby providing accurate and reliable position information for the maintenance and repair of the hot-spot cells to a certain extent, and further improving the convenience of the maintenance and repair of the hot-spot defective components.

[0099] In addition, as described above, in this embodiment, two different color modes are used to render the infrared panoramic image, which can respectively highlight the hot spot phenomena caused by different reasons. In the iron red color image, hot spots in the shapes of points, stripes, flocs, and planes can be highlighted; this type of hot spot is mostly caused by the hot spot effect due to faults in the photovoltaic cells or circuit faults in the photovoltaic module. For this type of fault, the general treatment method is that when the number of hot spot cells in a photovoltaic module reaches a certain amount, the photovoltaic module is disassembled and replaced. In the thermal color image, the hot spot phenomena caused by being blocked by dust, vegetation, etc. can be highlighted; for the treatment method of this type of fault, only the photovoltaic module needs to be cleaned.

[0100] Thus, when the user calibrates the hot spot cells in the infrared panoramic images rendered and displayed in two different color modes, the user can further label the hot spot fault types corresponding to the hot spot defect components containing the hot spot cells, that is, label the non-blocking hot spot faults caused by faults in the internal cells or circuits of the photovoltaic module or the blocking hot spot faults caused by the blocking of the photovoltaic module. Based on this fault type, reliable data support can be provided for the subsequent inspection and maintenance personnel for maintenance processing.

[0101] Moreover, because the visible light panoramic image can more realistically display the real image of the photovoltaic module, when the user calibrates the fault type of the hot spot defect component, the user can further verify and determine the fault type of the component with hot spot cells by combining the visible light panoramic image, that is, comprehensively verify and determine the hot spot fault type in the photovoltaic module by combining the visible light panoramic image and the infrared panoramic image.

[0102] Based on the above discussion, in an optional implementation manner of this embodiment, after the user inputs the hot spot fault type corresponding to each hot spot cell according to the displayed imaging area, it further includes:

[0103] According to the number of hot spot cell strings in each hot spot defect component, the first fault category of the hot spot defect component is divided into single-string cell fault, two-string cell fault, and three-string cell fault;

[0104] According to the distribution positions of the hot spot cells in the hot spot cell string, the second fault category of each hot spot cell string is divided into dot-shaped fault, dot-strip-shaped fault, and strip-shaped fault.

[0105] It can be understood that the hot spot defect component in this embodiment refers to the photovoltaic module with hot spot cells; and generally, each photovoltaic module contains multiple strings of electrically connected photovoltaic cell strings; the hot spot cell string is the photovoltaic cell string with hot spot cells in the hot spot defect component.

[0106] Refer to Figure 2, taking the example that each photovoltaic module in a photovoltaic power station contains three strings of photovoltaic cells connected in parallel, based on the working mode of each photovoltaic cell string in the photovoltaic module, it can be known that the photovoltaic cells in the same string of photovoltaic cells are connected in series with each other. As long as there is a fault in one photovoltaic cell in the same string of photovoltaic cells, it will directly affect the normal output of the current of the entire string of photovoltaic cells, that is, it is equivalent to the entire string of photovoltaic cells not working. Therefore, in this embodiment, the number of photovoltaic cell strings with hot spot cells in the photovoltaic module is used to divide the fault categories of the photovoltaic strings.

[0107] Such as Figure 2 shown, in this embodiment, the hot spot defect modules can be first divided into three major first fault categories according to the distribution of hot spot cells: single cell string fault, double cell string fault, and triple cell string fault, etc. It can be understood that a single cell string fault means that there is only one string of photovoltaic cells with hot spot cells in the photovoltaic module; a double cell string fault means that there are two strings of photovoltaic cells with hot spot cells in the photovoltaic module; a triple cell string fault means that all three strings of photovoltaic cells in the photovoltaic module have hot spot cells. In Figure 2 the shown embodiment, each small rectangular frame represents a photovoltaic cell, and the blackened dark rectangular frame is the hot spot cell.

[0108] On the basis of dividing the hot spot defect modules into the first fault categories according to the above-mentioned division principle of the first fault categories, the second fault categories are further divided; for each hot spot cell string in the hot spot defect module, it can be further divided into three second fault categories: dot fault, dot-strip fault, and strip fault; among them, a dot fault means that the hot spot cells in the same hot spot cell string are scattered individually; a strip fault means that the hot spot cells in the same hot spot cell string are continuously adjacent; a dot-strip fault means that the same hot spot cell string contains both scattered hot spot cells and multiple continuously adjacent hot spot cells.

[0109] Based on the above classification method, as Figure 3 shown, the fault category marks can be further made for each hot spot defect module. Among them, the position of a single cell string fault is marked as 01, the position of a double cell string fault is marked as 02, and the position of a triple cell string fault is marked as 03; in a single cell string fault, a double cell string fault, and a triple cell string fault, the hot spot shape corresponding to a dot fault is all marked as A, the hot spot shape corresponding to a dot-strip fault is all marked as B, and the hot spot shape corresponding to a strip fault is all marked as C; of course, it should be noted that for a triple cell string fault and each string of hot spot cell strings belongs to a strip fault, there must be a large area of the hot spot cell string with hot spot phenomena, so the strip fault in the triple cell string fault can be marked as a planar hot spot shape C.

[0110] Based on the above principles for fault category division, each hot spot defective component is numbered from 01 - n according to the marking order. Taking the punctiform hot spot of a single cell string as an example, it can be numbered as 01A01, 01A02. Among them, the first 01 indicates that it is a single cell string fault, A indicates that it is a punctiform fault, and the last 01 and 02 respectively represent the first and second hot spot defective components of the punctiform hot spot fault of the single cell string.

[0111] Through the above fault category division, the fault conditions of each hot spot defective component can be intuitively and detailedly displayed, so as to provide reliable data on the severity of the faults of each hot spot defective component, which helps the operation and maintenance personnel to determine the priority order of photovoltaic component repair, and thus assist the operation and maintenance personnel to formulate a more reasonable maintenance strategy.

[0112] S4: According to the geographical location information of the imaging position of the hot spot cell in the infrared panoramic image, display and output the distribution position of the hot spot cell in the visible light panoramic image and the hot spot fault type.

[0113] Based on the above discussion, while outputting the position information and hot spot fault type (i.e., occluded hot spot fault and non - occluded hot spot fault) of the hot spot defective component, the fault category of each hot spot defective component can also be output simultaneously. Thus, the operation and maintenance personnel can quickly determine a more reasonable maintenance strategy based on the position information, hot spot fault type, and fault category of the hot spot defective component; realizing simple, fast, and efficient maintenance of the photovoltaic power station.

[0114] In summary, in this application, an unmanned aerial vehicle is used to simultaneously scan and obtain the infrared panoramic image and visible light panoramic image of the photovoltaic power station; on this basis, the infrared panoramic image carrying temperature parameters is respectively rendered and displayed in two different rendering modes; thus, the user can directly and more simply and quickly identify and calibrate the hot spot cells that generate hot spot phenomena due to different reasons based on the images rendered in the two different rendering modes, so as to accurately and comprehensively identify the existing hot spot cells. After determining the hot spot cells, further use the visible light panoramic image to locate the position where the hot spot cells are located, thereby reducing the difficulty of maintaining and overhauling the hot spot cells. It can be seen that this application realizes simple and fast operation and maintenance of the photovoltaic power station and reduces the operation and maintenance cost of the power station.

[0115] Next, the hot spot fault detection device for photovoltaic components provided by the embodiments of the present invention will be introduced. The hot spot fault detection device for photovoltaic components described below can be correspondingly referred to the hot spot fault detection method for photovoltaic components described above.

[0116] Figure 4 is the structural block diagram of the hot spot fault detection device for photovoltaic components provided by the embodiments of the present invention. Refer to Figure 4 The hot spot fault detection device for photovoltaic components may include:

[0117] The image processing module 100 is configured to generate an infrared panoramic image and a visible light panoramic image respectively according to the infrared image and the visible light image obtained by the drone for dual-light image acquisition of the photovoltaic power station.

[0118] The rendering and display module 200 is configured to perform ferric color mode rendering and thermal color mode rendering on the infrared panoramic image respectively according to the temperature information data corresponding to each pixel point in the infrared panoramic image, generate and display a ferric color image and a thermal color image, so that the user can calibrate the hot spot battery through the ferric color image and the thermal color image.

[0119] The fault calibration module 300 is configured to calibrate the imaging position of the hot spot battery and the type of hot spot fault according to the ferric color image and the thermal color image.

[0120] The information output module 400 is configured to display and output the distribution position of the hot spot battery in the visible light panoramic image and the type of hot spot fault according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image.

[0121] In an optional embodiment of the present application, the image processing module 100 includes an infrared processing unit, which is configured to generate a raw file containing temperature information data according to the infrared image; perform numerical calculation conversion on the temperature information data in the raw file to obtain the temperature parameter corresponding to the infrared image; extract the geographical location information corresponding to the infrared image according to the position and attitude information of the drone when shooting each infrared image; convert each infrared image into a TIFF format image with the temperature parameter and geographical location information according to the geographical location information; and splice the TIFF format images to obtain an infrared panoramic image.

[0122] In an optional embodiment of the present application, the rendering and display module 200 is specifically configured to form a temperature parameter set with the temperature parameters of each pixel point in the infrared panoramic image as set elements; divide each temperature parameter element in the temperature parameter set into multiple temperature grading sets; use the natural breakpoint method formula , in combination with each temperature parameter element in the temperature parameter set, for and perform iterative optimization adjustment to obtain multiple temperature grading sets obtained by dividing each temperature parameter element when the variance goodness of fit is the largest and not less than the set goodness of fit threshold; where is the variance goodness of fit, is the th temperature parameter element in the temperature parameter set; is the average value of each temperature parameter element in the temperature parameter set. is the number of the temperature classification sets; is the number of the temperature parameter elements in the th temperature classification set; is the th temperature parameter element in the th temperature classification set; and

[0123] In an optional embodiment of the present application, the image processing module 100 is specifically configured to splice the visible light image through the Context Capture software to form the visible light panoramic image; wherein, the height at which the drone collects the visible light image and the infrared image is not higher than 40 m, and the overlap rate between two adjacent visible light images spliced along the heading or side direction of the drone is not less than 70%.

[0124] In an optional embodiment of the present application, the rendering and display module 200 is specifically configured to, after separately generating the visible light panoramic image and the infrared panoramic image, jointly import the visible light panoramic image and the infrared panoramic image into the LocaSpace Viewer software, and determine the corresponding relationship between the pixel points in the visible light panoramic image and the pixel points in the infrared panoramic image according to the geographical location information respectively corresponding to the visible light panoramic image and the infrared panoramic image; correspondingly, according to the geographical location information of the hot spot battery in the infrared panoramic image and the corresponding relationship, display and output the distribution position and the hot spot fault type of the hot spot battery in the visible light panoramic image.

[0125] In an optional embodiment of the present application, the information output module 400 is further configured to, after calibrating the imaging position and the hot spot fault type of the hot spot battery according to the iron red color image and the thermal color image, divide the first fault category of the hot spot defective component into a single-string battery fault, a two-string battery fault, and a three-string battery fault according to the number of hot spot battery strings in each of the hot spot defective components; divide the second fault category of each of the hot spot battery strings into a dot fault, a dot-strip fault, and a strip fault according to the distribution positions of the hot spot batteries in the hot spot battery string; and correspondingly, output the first fault category, the second fault type, the distribution position, and the hot spot fault type corresponding to each of the hot spot defective components; wherein the hot spot fault type includes an occlusion-type hot spot fault and a non-occlusion-type hot spot fault.

[0126] The fault detection device for a photovoltaic module in this embodiment is used to implement the foregoing fault detection method for a photovoltaic module. Therefore, the specific implementation manners in the fault detection device for a photovoltaic module can be seen in the embodiment part of the fault detection method for a photovoltaic module in the foregoing text. The specific implementation manners can refer to the descriptions of the corresponding individual part embodiments and will not be elaborated herein.

[0127] An embodiment of a fault detection device for a photovoltaic module is also provided in the present application. The fault detection device for a photovoltaic module may include:

[0128] A memory for storing a computer program;

[0129] A processor for executing the computer program to implement the steps of the fault detection method for a photovoltaic module as described in any one of the foregoing.

[0130] The steps of the fault detection method for a photovoltaic module executed by the processor include:

[0131] Performing dual-light image acquisition on a photovoltaic power station by using a drone to obtain an infrared image and a visible light image, and respectively generating an infrared panoramic image and a visible light panoramic image;

[0132] Performing iron red color mode rendering and thermal color mode rendering on the infrared panoramic image respectively according to the temperature parameters corresponding to the pixel points in the infrared panoramic image to generate an iron red color image and a thermal color image;

[0133] Calibrating the imaging position and the hot spot fault type of the hot spot battery according to the iron red color image and the thermal color image;

[0134] Displaying and outputting the distribution position and the hot spot fault type of the hot spot battery in the visible light panoramic image according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image.

[0135] The present application also provides a computer-readable storage medium storing a computer program, which when executed, implements the steps of the fault detection method for a photovoltaic module as described in any one of the above.

[0136] The computer-readable storage medium may be a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0137] It should be noted that in this text, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes the inherent elements thereof. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device that includes the said element. In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the corresponding technical solutions in the prior art in terms of implementation principles are not described in detail to avoid unnecessary repetition.

[0138] Specific examples are used in this text to elaborate on the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention. It should be pointed out that for those of ordinary skill in the art in this technical field, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the present invention.

Claims

1. A method for detecting hot spot faults of photovoltaic modules, characterized in that: include: The infrared image and the visible light image are obtained by collecting dual-light images of the photovoltaic power station by the drone, and an infrared panoramic image and a visible light panoramic image are generated respectively; According to the temperature parameters corresponding to each pixel point in the infrared panoramic image, the infrared panoramic image is rendered in an iron red color mode and a thermal color mode to generate an iron red color image and a thermal color image; Calibrate the imaging position of the hot spot battery and the hot spot fault type according to the iron red color image and the thermal color image; According to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image, the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type are displayed and output.

2. The method for detecting hot spot failure of a photovoltaic module according to claim 1, characterized in that: The process of generating the infrared panoramic image includes: Generate a raw file containing temperature information data according to the infrared image; Performing numerical calculation conversion on the temperature information data in the raw file to obtain the temperature parameters corresponding to the infrared image; Extracting geographical location information corresponding to each infrared image according to the position and posture information when the drone takes each infrared image; Convert each of the infrared images into a TIFF format image with the temperature parameter and the geographical location information according to the geographical location information; The TIFF format images are stitched together to obtain the infrared panoramic image.

3. The method for detecting hot spot faults of photovoltaic modules according to claim 1, characterized in that: The process of rendering the infrared panoramic image according to the temperature information data of the infrared panoramic image to generate an iron red color image and a thermal color image includes: Taking the temperature parameter of each pixel in the infrared panoramic image as a set element, forming a temperature parameter set; Dividing each temperature parameter element in the temperature parameter set into a plurality of temperature classification sets; Using the natural breakpoint method formula , combined with each temperature parameter element in the temperature parameter set, and Iterative optimization adjustment is performed to obtain a plurality of temperature classification sets for dividing each temperature parameter element when the variance goodness of fit is the largest and not less than a set goodness of fit threshold; wherein, is the variance goodness of fit, is the first temperature parameter in the set Temperature parameter elements; is the average value of each temperature parameter element in the temperature parameter set; the number of sets of temperature classifications; For the the number of the temperature parameter elements in the temperature classification set; For the The first temperature classification set The temperature parameter element, For the The average value of each of the temperature parameter elements in the temperature classification set; and , ; Taking the maximum temperature parameter element and the minimum temperature parameter in each of the temperature classification sets as the primary color temperature parameter interval; Adopting the iron red color mode, for the pixel points whose corresponding temperature parameter elements belong to the color temperature parameter interval of the same level in the infrared panoramic image, the same color parameter is assigned, and the pixel points whose corresponding temperature parameter elements belong to the color temperature parameter interval of different levels are assigned different color parameters for rendering, so as to obtain the iron red color rendering image; The thermal color mode is adopted. For the pixel points whose corresponding temperature parameter elements belong to the color temperature parameter interval of the same level in the infrared panoramic image are assigned the same color parameter, and the pixel points whose corresponding temperature parameter elements belong to the color temperature parameter interval of different levels are assigned different color parameters for rendering to obtain the thermal color rendering image.

4. The method for detecting hot spot failure of a photovoltaic module according to claim 1, characterized in that: The process of generating a visible light panoramic image includes: The visible light images are stitched using Context Capture software to form the visible light panoramic image; wherein the height at which the drone collects the visible light image and the infrared image is not higher than 40 m, and the overlap rate between the two stitched visible light images along the heading or sideways of the drone is not less than 70%.

5. The method for detecting hot spot failure of a photovoltaic module according to claim 1, characterized in that: After generating the visible light panoramic image and the infrared panoramic image respectively, it also includes: Importing the visible light panoramic image and the infrared panoramic image into LocaSpace Viewer software, and determining the correspondence between pixel points in the visible light panoramic image and pixel points in the infrared panoramic image according to the geographical location information corresponding to the visible light panoramic image and the infrared panoramic image respectively; Accordingly, according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image, the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type are displayed and output, including: According to the geographical location information of the hot spot battery in the infrared panoramic image and the corresponding relationship, the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type are displayed and output.

6. The photovoltaic module fault detection method according to any one of claims 1 to 5, characterized in that: After calibrating the imaging position of the hot spot battery and the hot spot fault type according to the iron red color image and the thermal color image, the method further includes: According to the number of hot spot battery strings in each of the hot spot defective components, the first fault category of the hot spot defective components is divided into a single-string battery failure, a two-string battery failure and a three-string battery failure; According to the distribution position of each of the hot spot batteries in the hot spot battery string, the second fault category of each of the hot spot battery strings is divided into point fault, point strip fault and strip fault; Accordingly, displaying and outputting the distribution position of the hot spot cells in the visible light panoramic image and the hot spot fault type includes: Output the first fault category, the second fault type, the distribution location and the hot spot fault type corresponding to each of the hot spot defective components; wherein the hot spot fault type includes an obstruction type hot spot fault and a non-obstruction type hot spot fault.

7. A photovoltaic module fault detection device, characterized in that: include: An image processing module is used to obtain an infrared image and a visible light image according to the dual-light image acquisition of the photovoltaic power station by the drone, and generate an infrared panoramic image and a visible light panoramic image respectively; A rendering and display module, used to perform iron red color mode rendering and thermal color mode rendering on the infrared panoramic image according to the temperature information data corresponding to each pixel point in the infrared panoramic image, so as to generate an iron red color image and a thermal color image; A fault calibration module, used to calibrate the imaging position of the hot spot battery and the hot spot fault type according to the iron red color image and the thermal color image; The information output module is used to display and output the distribution position of the hot spot battery in the visible light panoramic image and the hot spot fault type according to the geographical location information of the imaging position of the hot spot battery in the infrared panoramic image.

8. The photovoltaic module fault detection device according to claim 7, characterized in that: The image processing module includes an infrared processing unit, which is used to generate a raw file containing temperature information data based on the infrared image; perform numerical calculation conversion on the temperature information data in the raw file to obtain temperature parameters corresponding to the infrared image; extract geographical location information corresponding to the infrared image based on the position and posture information when the drone takes each infrared image; convert each infrared image into a TIFF format image with the temperature parameters and geographical location information based on the geographical location information; and perform image stitching on each TIFF format image to obtain an infrared panoramic image.

9. A photovoltaic module fault detection device, characterized in that: include: Memory for storing computer programs; A processor is used to execute the computer program to implement the steps of the photovoltaic component fault detection method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed to implement the steps of the photovoltaic component fault detection method according to any one of claims 1 to 6.

Images