A photovoltaic panel self-adaptive automatic cleaning method and device, electronic equipment and product

Abstract

This invention discloses an adaptive automatic cleaning method, device, electronic device, and product for photovoltaic panels, relating to the field of computer vision technology. The method involves extracting visible light and thermal images of each photovoltaic panel; converting the visible light images into grayscale images; performing binary segmentation on the corresponding grayscale images to obtain the dust-covered areas of each photovoltaic panel; determining the average temperature of the dust-covered areas of each photovoltaic panel based on the thermal images; determining the dust coverage of the dust-covered areas of each photovoltaic panel based on the average temperature and average ambient temperature; determining the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered areas, and determining the set of photovoltaic panels to be cleaned based on the comprehensive dust coverage; iteratively optimizing the cleaning path for the set of photovoltaic panels to be cleaned to obtain the final cleaning path, and controlling a cleaning robot to clean the photovoltaic panels to be cleaned. This invention can promptly detect and clean photovoltaic panels when dust coverage reaches a certain level.

Description

Technical Field

[0001] This invention belongs to the field of computer vision technology, specifically relating to an adaptive automatic cleaning method, device, electronic equipment, and product for photovoltaic panels. Background Technology

[0002] Solar photovoltaic (PV) power generation, as a clean and renewable energy source, has been widely adopted globally in recent years, especially in solar-rich regions such as western my country, where PV power plants have developed rapidly. However, these regions are characterized by dry climates and frequent dust storms, making it easy for dust to accumulate on the surface of PV modules. As dust accumulates, the light transmittance of the PV panels decreases significantly, leading to a substantial drop in power generation efficiency. In severe cases, this can even trigger hot spot effects, accelerate module aging, and shorten the overall lifespan of the power plant. Therefore, timely and effective cleaning of PV panels is a crucial step in ensuring the efficient and stable operation of PV power plants.

[0003] Currently, the common cleaning strategy for photovoltaic power plants is periodic cleaning at fixed intervals, such as a comprehensive cleaning once a month or quarter. However, due to the differences in the rate of dust accumulation in different areas—the dust accumulation rate in areas near wind vents, near the ground, or adjacent to roads is much faster than in other areas—fixed-cycle cleaning either leads to resource waste due to cleaning too early or to inefficient power generation for a long period of time due to cleaning too late, resulting in unnecessary power generation losses.

[0004] Therefore, how to provide an effective solution to promptly detect and clean photovoltaic panels when they are covered with dust to a certain extent in photovoltaic power plants has become an urgent problem to be solved in the existing technology. Summary of the Invention

[0005] The purpose of this invention is to provide a method, apparatus, electronic device, and product for adaptive automatic cleaning of photovoltaic panels, in order to solve the above-mentioned problems existing in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides an adaptive automatic cleaning method for photovoltaic panels, comprising: Receive regional visible light images and regional thermal images of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera; The visible light images of each photovoltaic panel in the visible light image of the region and the thermal imaging images of each photovoltaic panel in the thermal imaging image of the region are extracted by the target detection algorithm. The visible light images of each photovoltaic panel are converted into grayscale images, and the grayscale histogram of the corresponding grayscale image of each photovoltaic panel is extracted. For each photovoltaic panel's grayscale histogram, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the preset grayscale range is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel. Based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, the grayscale image corresponding to each photovoltaic panel is binary segmented, and the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel are taken as the pixels of the dust-covered area, thus obtaining the dust-covered area corresponding to each photovoltaic panel. Based on the thermal imaging images of each photovoltaic panel, the average temperature of the dust-covered area corresponding to each photovoltaic panel is determined. The dust coverage of the dust-covered area corresponding to each photovoltaic panel is determined based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current period. Based on the area and dust coverage of each photovoltaic panel, the overall dust coverage of each photovoltaic panel is determined. Photovoltaic panels whose overall dust coverage exceeds a preset threshold are identified as photovoltaic panels to be cleaned, thus obtaining a set of photovoltaic panels to be cleaned. With the goal of minimizing the path distance, the cleaning path of the set of photovoltaic panels to be cleaned is iteratively optimized using a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned. The cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned. The cleaning robot is controlled to clean each photovoltaic panel along the final cleaning path.

[0007] In one possible design, after converting the visible light images of each photovoltaic panel into grayscale images, the method further includes: Filter the grayscale image corresponding to each photovoltaic panel; Morphological processing is performed on the filtered grayscale images corresponding to each photovoltaic panel to obtain the processed grayscale images corresponding to each photovoltaic panel. The step of extracting the grayscale histogram of the grayscale image corresponding to each photovoltaic panel includes: Extract the grayscale histogram of the processed grayscale image corresponding to each photovoltaic panel.

[0008] In one possible design, for any grayscale histogram corresponding to a photovoltaic panel, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within a preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel, including: For any photovoltaic panel, determine whether there are two frequency peaks in the gray-level histogram of any photovoltaic panel within the first preset gray-level interval; If the grayscale histogram corresponding to any photovoltaic panel contains two frequency peaks within the first preset grayscale interval, then the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the first preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel. If the grayscale histogram corresponding to any photovoltaic panel does not have two frequency peaks within the first preset grayscale interval, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the second preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel, wherein the first preset grayscale interval is a subset of the second preset grayscale interval.

[0009] In one possible design, the goal is to minimize the path distance, and a genetic algorithm is used to iteratively optimize the cleaning paths for the set of photovoltaic panels to be cleaned, resulting in the final cleaning path for each panel, including: After encoding each photovoltaic panel in the set of photovoltaic panels to be cleaned, a population is initialized, and each chromosome in the population represents a cleaning path; The fitness value of each chromosome in the population is calculated based on the path distances corresponding to each chromosome in the population. The population is updated iteratively by selecting the chromosomes with the highest fitness values ​​from the population through tournament selection and selecting individuals from the remaining chromosomes of the population through roulette wheel selection for crossover and mutation operations. When the iteration stops, the cleaning path corresponding to the chromosome with the highest fitness value is taken as the final cleaning path.

[0010] In one possible design, the fitness value of the chromosome is... ,in, , , This represents the fitness value of a specific chromosome in the k-th generation. This represents the maximum path distance among all chromosomes in the k-th generation. This represents the path distance corresponding to a specific chromosome in the k-th generation. This represents the selection pressure regulation value of the chromosome in the kth generation. Let M and c represent the selection pressure regulation value of the chromosome in the (k-1)th generation. M and c are both constants, and c ∈ [0.95, 0.99].

[0011] In one possible design, multiple chromosomes with the highest fitness values ​​are selected from the population using tournament selection as offspring chromosomes, and individuals are selected from the remaining chromosomes in the population using roulette wheel selection for crossover and mutation operations to iteratively update the population, including: The tournament selection method is used to select the chromosomes with the highest fitness values ​​from the population as offspring chromosomes. Randomly select N chromosomes from the remaining chromosomes in the population; The offspring chromosome is obtained by selecting one chromosome from the N chromosomes using a roulette wheel selection method as the parent chromosome and performing crossover and mutation operations. N chromosomes are randomly selected from the remaining chromosomes of the population. Then, the parent chromosomes are selected from the randomly selected N chromosomes using a roulette wheel selection method for crossover and mutation operations. When the total number of offspring chromosomes reaches the preset number, the population iteration update is considered complete, and all offspring chromosomes are updated to parent chromosomes for use in the next population iteration update.

[0012] In one possible design, a chromosome is selected from the N chromosomes using a roulette wheel selection method as the parent chromosome for crossover and mutation operations to obtain offspring chromosomes, including: A chromosome is selected from the N chromosomes using a roulette wheel selection method as the first parent chromosome; Determine whether the path distance corresponding to the first parent chromosome is greater than a distance threshold, wherein the distance threshold is determined based on the number of photovoltaic panels to be cleaned in the set of photovoltaic panels to be cleaned and the average distance between the photovoltaic panels to be cleaned in the set of photovoltaic panels to be cleaned; If the path distance corresponding to the first parent chromosome is greater than the distance threshold, then a chromosome is randomly selected from the N chromosomes as the second parent chromosome, and the first parent chromosome and the second parent chromosome are sequentially crossed and then mutated according to the preset mutation probability to obtain the offspring chromosome; If the path distance corresponding to the first parent chromosome is less than or equal to the distance threshold, then the first parent chromosome undergoes a crossover operation and a mutation operation is performed according to a preset mutation probability to obtain the offspring chromosome.

[0013] In a second aspect, the present invention provides a photovoltaic panel adaptive automatic cleaning device, comprising: The receiving unit is used to receive the regional visible light image and regional thermal imaging image of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera; The extraction unit is used to extract the visible light images of each photovoltaic panel in the visible light image of the region and the thermal imaging images of each photovoltaic panel in the thermal imaging image of the region using a target detection algorithm. The conversion unit is used to convert the visible light image of each photovoltaic panel into a grayscale image; The extraction unit is also used to extract the grayscale histogram of the grayscale image corresponding to each photovoltaic panel; The segmentation threshold selection unit is used to select the gray value corresponding to the minimum frequency between the two maximum frequencies within a preset gray range as the binary segmentation threshold of the gray-level image corresponding to the photovoltaic panel for the gray-level histogram corresponding to each photovoltaic panel. The binary segmentation unit is used to perform binary segmentation on the grayscale image corresponding to each photovoltaic panel based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, and to take the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel as the pixels of the dust-covered area, so as to obtain the dust-covered area corresponding to each photovoltaic panel. The determination unit is used to determine the average temperature of the dust-covered area corresponding to each photovoltaic panel based on the thermal imaging images of each photovoltaic panel. The determining unit is also used to determine the dust coverage of the dust-covered area corresponding to each photovoltaic panel based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current time period. The determining unit is also used to determine the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered area corresponding to each photovoltaic panel. A photovoltaic panel selection unit is used to select photovoltaic panels whose overall dust coverage exceeds a preset threshold as photovoltaic panels to be cleaned, thereby obtaining a set of photovoltaic panels to be cleaned. The iterative optimization unit is used to minimize the path distance and iteratively optimize the cleaning path of the set of photovoltaic panels to be cleaned using a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned. The cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned. The cleaning control unit is used to control the cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path.

[0014] Thirdly, the present invention provides an electronic device comprising a memory, a processor, and a transceiver connected in sequence and communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the photovoltaic panel adaptive automatic cleaning method as described in the first aspect or any possible design of the first aspect.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, perform the photovoltaic panel adaptive automatic cleaning method described in the first aspect or any possible design of the first aspect.

[0016] Fifthly, the present invention provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the photovoltaic panel adaptive automatic cleaning method as described in the first aspect or any possible design of the first aspect.

[0017] Beneficial effects: This invention discloses an adaptive automatic cleaning scheme for photovoltaic panels, which can promptly detect when the dust coverage of photovoltaic panels in a photovoltaic power station reaches a certain level, and control a cleaning robot to clean the photovoltaic panels using the shortest path, thereby improving the overall power generation efficiency and economy of the photovoltaic power station. Specifically, the system receives regional visible light images and regional thermal images of the photovoltaic panel area in a photovoltaic power station acquired by a dual-light camera; extracts the visible light images of each photovoltaic panel in the regional visible light images and the thermal images of each photovoltaic panel in the regional thermal images using a target detection algorithm; converts the visible light images of each photovoltaic panel into grayscale images and extracts the grayscale histogram of the corresponding grayscale image of each photovoltaic panel; for the grayscale histogram of each photovoltaic panel, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within a preset grayscale range is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel; based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, binary segmentation is performed on the grayscale image corresponding to each photovoltaic panel, and the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel are taken as the pixels of the dust-covered area, thus obtaining the grayscale image corresponding to each photovoltaic panel. The process involves identifying dust-covered areas; determining the average temperature of the dust-covered area corresponding to each photovoltaic panel based on thermal imaging images; determining the dust coverage of the dust-covered area corresponding to each photovoltaic panel based on the average temperature of the dust-covered area and the average ambient temperature during the current time period; determining the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered area corresponding to each photovoltaic panel; identifying photovoltaic panels with a comprehensive dust coverage exceeding a preset threshold as photovoltaic panels to be cleaned, thus obtaining a set of photovoltaic panels to be cleaned; iteratively optimizing the cleaning path of the set of photovoltaic panels to be cleaned using a genetic algorithm with the goal of minimizing path distance, thus obtaining the final cleaning path for each photovoltaic panel to be cleaned, wherein the cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned; and controlling a cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path. In this way, the dust-covered area can be predicted by histogram and the dust coverage of the dust-covered area can be judged by thermal imaging. This allows for the rapid and accurate determination of the overall dust coverage of each photovoltaic panel. When the overall dust coverage of the photovoltaic panels in the photovoltaic power station reaches a certain level, the shortest path for cleaning the photovoltaic panels can be planned. The cleaning robot can then be controlled to clean the photovoltaic panels using the shortest path. This achieves automatic detection of dust coverage on photovoltaic panels and enables the cleaning robot to complete the cleaning of the photovoltaic panels in the shortest time by controlling the shortest path. This minimizes the impact of dust coverage on the overall power generation efficiency of the photovoltaic power station, improves the overall power generation efficiency and economy of the photovoltaic power station, and facilitates practical application and promotion. Attached Figure Description

[0018] Figure 1 A flowchart of the photovoltaic panel adaptive automatic cleaning method provided in the embodiments of this application; Figure 2 A block diagram of the photovoltaic panel adaptive automatic cleaning device provided in the embodiments of this application; Figure 3 This is a block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0020] It should be understood that although the terms first, second, etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, without departing from the scope of the exemplary embodiments of the invention.

[0021] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0022] It should be understood that specific details are provided in the following description to facilitate a complete understanding of the exemplary embodiments. However, those skilled in the art will understand that the exemplary embodiments can be implemented without these specific details. For example, the system may be shown in block diagrams to avoid obscuring the example with unnecessary details. In other instances, well-known processes, structures, and techniques may be shown without unnecessary details to avoid obscuring the exemplary embodiments.

[0023] In order to promptly detect and clean photovoltaic panels when dust coverage reaches a certain level in photovoltaic power plants, this application provides an adaptive automatic cleaning method, device, electronic device, and product for photovoltaic panels. This adaptive automatic cleaning method, device, electronic device, and product can promptly detect when dust coverage on photovoltaic panels in photovoltaic power plants reaches a certain level, and control a cleaning robot to clean the photovoltaic panels using the shortest path, thereby improving the overall power generation efficiency and economy of photovoltaic power plants.

[0024] The adaptive automatic cleaning method for photovoltaic panels provided in this application can be applied to the back-end management terminal or server in a photovoltaic power station. It is understood that the execution entity described herein does not constitute a limitation on the embodiments of this application.

[0025] The adaptive automatic cleaning method for photovoltaic panels provided in the embodiments of this application will be described in detail below.

[0026] like Figure 1 The diagram shown is a flowchart of a photovoltaic panel adaptive automatic cleaning method provided in the first aspect of the present application. The photovoltaic panel adaptive automatic cleaning method may include, but is not limited to, the following steps S101-S111.

[0027] Step S101. Receive the visible light image and thermal imaging image of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera.

[0028] Among them, the dual-light camera is a camera that integrates a visible light lens and an infrared thermal imaging lens. It can be used to capture high-resolution color images under visible light, and at the same time, it can receive infrared radiation emitted by objects to form a thermal imaging image that shows the temperature distribution.

[0029] Depending on the distribution of photovoltaic (PV) panels in a photovoltaic (PV) power station, one or more dual-light cameras can be installed to ensure coverage of all PV panel areas. During operation, the dual-light cameras can acquire and upload real-time visible light and thermal images of the PV panel areas within their field of view. Each PV panel area contains multiple PV panels.

[0030] Step S102. Extract the visible light images of each photovoltaic panel in the regional visible light image and the thermal imaging images of each photovoltaic panel in the regional thermal imaging image using the target detection algorithm.

[0031] The target detection algorithm may be, but is not limited to, YOLO (You Only Look Once) algorithm, region-based convolutional neural network (R-CNN) algorithm, etc., and is not specifically limited in the embodiments of this application.

[0032] Step S103. Convert the visible light image of each photovoltaic panel into a grayscale image, and extract the grayscale histogram of the corresponding grayscale image of each photovoltaic panel.

[0033] Specifically, when extracting the grayscale histogram of the grayscale image corresponding to each photovoltaic panel, the number of pixels of each gray level appearing in the image can be counted to obtain the grayscale histogram of the grayscale image corresponding to each photovoltaic panel.

[0034] In one or more embodiments, after converting the visible light image of each photovoltaic panel into a grayscale image, the grayscale image corresponding to each photovoltaic panel can be filtered, and then morphological processing (i.e., dilation and erosion) can be performed on the filtered grayscale image corresponding to each photovoltaic panel to obtain the processed grayscale image corresponding to each photovoltaic panel. Finally, the grayscale histogram of the processed grayscale image corresponding to each photovoltaic panel is extracted. Filtering the grayscale image can reduce the interference of noise on subsequent steps, and morphological processing can remove noise points in the image.

[0035] Step S104. For the grayscale histogram corresponding to each photovoltaic panel, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the preset grayscale range is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel.

[0036] For any grayscale histogram corresponding to a photovoltaic panel, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the preset grayscale range is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel. This may include, but is not limited to, the following steps S1041-S1043.

[0037] Step S1041. For any photovoltaic panel, determine whether there are two frequency peaks in the gray-level histogram of any photovoltaic panel within the first preset gray-level interval.

[0038] In a grayscale histogram, frequency refers to the number of pixels in the image that have a specific grayscale value. The first preset grayscale range can be set based on historical experience; for example, the first preset grayscale range can be 80-200.

[0039] Step S1042. If the grayscale histogram corresponding to any photovoltaic panel has two frequency peaks within the first preset grayscale interval, then the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the first preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel.

[0040] Step S1043. If the grayscale histogram corresponding to any photovoltaic panel does not have two frequency peaks in the first preset grayscale interval, the grayscale value corresponding to the minimum frequency between the two maximum frequencies in the second preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel.

[0041] The second preset grayscale range can also be set based on historical experience, and the first preset grayscale range is a subset of the second preset grayscale range. For example, if the first preset grayscale range is 80-200, then the second preset grayscale range can be 50-230.

[0042] Understandably, if the grayscale histogram corresponding to any photovoltaic panel still does not contain two frequency peaks within the second preset grayscale range, then the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the global grayscale range (0-255) can be used as the binary segmentation threshold for the grayscale image corresponding to any photovoltaic panel. Alternatively, the binary segmentation threshold of the grayscale image corresponding to a nearby photovoltaic panel can be used as the binary segmentation threshold for the grayscale image corresponding to any photovoltaic panel.

[0043] Step S105. Based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, perform binary segmentation on the grayscale image corresponding to each photovoltaic panel, and take the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel as the pixels of the dust-covered area, thereby obtaining the dust-covered area corresponding to each photovoltaic panel.

[0044] In one or more embodiments, when performing binary segmentation on the grayscale image corresponding to each photovoltaic panel, edge pixels of the grayscale image corresponding to each photovoltaic panel can be removed to avoid interference from the edge lines of the photovoltaic panels on the detection of dust-covered areas.

[0045] Step S106. Based on the thermal imaging images of each photovoltaic panel, determine the average temperature of the dust-covered area corresponding to each photovoltaic panel.

[0046] Specifically, the average temperature of the dust-covered area can be obtained by averaging the temperatures of multiple sampling points in the thermal imaging image of each photovoltaic panel and the area corresponding to the dust-covered area.

[0047] Step S107. Determine the dust coverage of the dust-covered area corresponding to each photovoltaic panel based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current time period.

[0048] When dust accumulates on photovoltaic (PV) panels, it triggers a hot spot effect, causing an abnormal increase in temperature in the dust-covered areas of the PV panel surface. The more severe the dust accumulation, the higher the temperature. Simultaneously, as the ambient temperature rises, the surface temperature of the PV panels also increases. Based on this, the ambient temperature can be divided into multiple temperature ranges, and the average temperature of the PV panel surface under different dust coverage levels can be statistically analyzed for each temperature range, resulting in a statistical table of PV panel surface temperatures under different dust coverage levels for each ambient temperature range. After determining the average temperature of the dust-covered area corresponding to each PV panel, the dust coverage level of the dust-covered area corresponding to each PV panel can be determined based on the statistical table, the average temperature of the dust-covered area corresponding to each PV panel, and the average ambient temperature during the current period.

[0049] Step S108. Based on the area and dust coverage of the dust-covered region corresponding to each photovoltaic panel, determine the comprehensive dust coverage of each photovoltaic panel.

[0050] In one or more embodiments, the dust coverage of the dust-covered area corresponding to each photovoltaic panel can be quantified and multiplied by the area of ​​the corresponding dust-covered area to obtain the comprehensive dust coverage of each photovoltaic panel.

[0051] Step S109. Select photovoltaic panels with a comprehensive dust coverage exceeding a preset threshold as photovoltaic panels to be cleaned, and obtain a set of photovoltaic panels to be cleaned.

[0052] The preset threshold can be set based on experience.

[0053] Step S110. With the goal of minimizing the path distance, the cleaning path of the set of photovoltaic panels to be cleaned is iteratively optimized using a genetic algorithm to obtain the final cleaning path for each photovoltaic panel to be cleaned.

[0054] The cleaning path records the cleaning order of each photovoltaic panel in the set of photovoltaic panels to be cleaned.

[0055] In one or more embodiments, with the goal of minimizing path distance, the cleaning path of the set of photovoltaic panels to be cleaned is iteratively optimized by a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned, which may include, but is not limited to, the following steps S1101-S1104.

[0056] Step S1101. After encoding each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned, initialize the population. Each chromosome in the population represents a cleaning path.

[0057] Step S1102. Calculate the fitness value of each chromosome in the population based on the path distance corresponding to each chromosome.

[0058] In one or more embodiments, the fitness value of a chromosome can be expressed as: ,in, , , This represents the fitness value of a specific chromosome in the k-th generation. This represents the maximum path distance among all chromosomes in the k-th generation. This represents the path distance corresponding to a specific chromosome in the k-th generation. This represents the selection pressure regulation value of the chromosome in the kth generation. This represents the selection pressure regulation value of the chromosome in the (k-1)th generation. M and c are both constants, and c∈[0.95,0.99]. M is a large value, which can be 300, 400 or 500, etc.

[0059] Step S1103. Select the chromosomes with the highest fitness values ​​from the population using tournament selection as offspring chromosomes, and select individuals from the remaining chromosomes of the population using roulette wheel selection to perform crossover and mutation operations, so as to iteratively update the population.

[0060] Specifically, a tournament selection method can be used to select the chromosomes with the highest fitness values ​​from the population as offspring chromosomes. Then, N chromosomes are randomly selected from the remaining chromosomes in the population, and a roulette wheel selection method is used to select one of these N chromosomes as the parent chromosome for crossover and mutation operations to obtain offspring chromosomes. Next, N chromosomes are randomly selected again from the remaining chromosomes in the population, and a roulette wheel selection method is used to select a parent chromosome from these N randomly selected chromosomes for crossover and mutation operations, until the total number of offspring chromosomes reaches a preset number. At this point, the current population iteration update is considered complete, and all offspring chromosomes are updated to parent chromosomes for use in the next population iteration update.

[0061] When selecting a chromosome from N chromosomes using the roulette wheel selection method as the parent chromosome for crossover and mutation operations, a first parent chromosome can be selected from the N chromosomes using the roulette wheel selection method. Then, it is determined whether the path distance corresponding to the first parent chromosome is greater than a distance threshold. This distance threshold is determined based on the number of photovoltaic panels to be cleaned in the set and the average distance between the photovoltaic panels to be cleaned in the set. If the path distance corresponding to the first parent chromosome is greater than the distance threshold, a second parent chromosome is randomly selected from the N chromosomes. The first parent chromosome and the second parent chromosome are then sequentially crossed, and mutation operations are performed according to a preset mutation probability to obtain offspring chromosomes. If the path distance corresponding to the first parent chromosome is less than or equal to the distance threshold, the first parent chromosome undergoes its own crossover operation, and mutation operations are performed according to a preset mutation probability to obtain offspring chromosomes.

[0062] By combining tournament selection and roulette wheel selection methods for iterative optimization of chromosome selection, the convergence speed of the algorithm can be improved. Simultaneously, sampling-selective sequential crossover and self-crossover operations make it easier for the algorithm to escape local optima, avoiding getting trapped in them and improving the algorithm's stability.

[0063] Step S1104. When the iteration stopping condition is met, the cleaning path corresponding to the chromosome with the highest fitness value is taken as the final cleaning path.

[0064] The iteration stopping condition can be reaching the maximum number of iterations or the convergence of the fitness value corresponding to the chromosome.

[0065] Step S111. Control the cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path.

[0066] The adaptive automatic cleaning method for photovoltaic panels provided by this invention involves receiving a visible light image and a thermal imaging image of a photovoltaic panel area in a photovoltaic power station acquired by a dual-light camera; extracting the visible light image of each photovoltaic panel from the visible light image and the thermal imaging image of each photovoltaic panel from the thermal imaging image using a target detection algorithm; converting the visible light image of each photovoltaic panel into a grayscale image and extracting the grayscale histogram of the corresponding grayscale image for each photovoltaic panel; using the grayscale histogram of each photovoltaic panel, taking the grayscale value corresponding to the minimum frequency between the two maximum frequencies within a preset grayscale range as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel; performing binary segmentation on the grayscale image corresponding to each photovoltaic panel based on the binary segmentation threshold, and taking the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel as the pixels of the dust-covered area, thus obtaining... The process involves identifying the dust-covered area corresponding to each photovoltaic panel; determining the average temperature of the dust-covered area based on the thermal imaging images of each photovoltaic panel; determining the dust coverage of the dust-covered area based on the average temperature of the dust-covered area and the average ambient temperature during the current time period; determining the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered area; identifying photovoltaic panels with a comprehensive dust coverage exceeding a preset threshold as the photovoltaic panels to be cleaned, thus obtaining a set of photovoltaic panels to be cleaned; iteratively optimizing the cleaning path of the set of photovoltaic panels to be cleaned using a genetic algorithm with the goal of minimizing the path distance, thus obtaining the final cleaning path for each photovoltaic panel to be cleaned, wherein the cleaning path records the cleaning order of each photovoltaic panel in the set of photovoltaic panels to be cleaned; and controlling a cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path. In this way, the dust-covered area can be predicted using histograms, and the dust coverage of the dust-covered area can be judged by combining thermal imaging. This allows for the rapid and accurate determination of the overall dust coverage of each photovoltaic panel. When the overall dust coverage of the photovoltaic panels in the photovoltaic power station reaches a certain level, the shortest path for cleaning the photovoltaic panels can be planned. The cleaning robot is then controlled to clean the photovoltaic panels using the shortest path. This achieves automatic detection of dust coverage on photovoltaic panels and enables the cleaning robot to complete the cleaning of the photovoltaic panels in the shortest time, minimizing the impact of dust coverage on the overall power generation efficiency of the photovoltaic power station, improving the overall power generation efficiency and economy, and facilitating practical application and promotion. Furthermore, during the optimization of the cleaning path, iterative optimization using a combination of tournament selection and roulette wheel selection improves the convergence speed of the algorithm. In addition, the sequential crossover and self-crossover operations of sampling selectivity make it easier for the algorithm to escape local optima and avoid getting trapped in them, thus improving the stability of the optimization algorithm.

[0067] Please see Figure 2The second aspect of this application provides an adaptive automatic cleaning device for photovoltaic panels, comprising: The receiving unit is used to receive the regional visible light image and regional thermal imaging image of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera; The extraction unit is used to extract the visible light images of each photovoltaic panel in the visible light image of the region and the thermal imaging images of each photovoltaic panel in the thermal imaging image of the region using a target detection algorithm. The conversion unit is used to convert the visible light image of each photovoltaic panel into a grayscale image; The extraction unit is also used to extract the grayscale histogram of the grayscale image corresponding to each photovoltaic panel; The segmentation threshold selection unit is used to select the gray value corresponding to the minimum frequency between the two maximum frequencies within a preset gray range as the binary segmentation threshold of the gray-level image corresponding to the photovoltaic panel for the gray-level histogram corresponding to each photovoltaic panel. The binary segmentation unit is used to perform binary segmentation on the grayscale image corresponding to each photovoltaic panel based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, and to take the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel as the pixels of the dust-covered area, so as to obtain the dust-covered area corresponding to each photovoltaic panel. The determination unit is used to determine the average temperature of the dust-covered area corresponding to each photovoltaic panel based on the thermal imaging images of each photovoltaic panel. The determining unit is also used to determine the dust coverage of the dust-covered area corresponding to each photovoltaic panel based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current time period. The determining unit is also used to determine the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered area corresponding to each photovoltaic panel. A photovoltaic panel selection unit is used to select photovoltaic panels whose overall dust coverage exceeds a preset threshold as photovoltaic panels to be cleaned, thereby obtaining a set of photovoltaic panels to be cleaned. The iterative optimization unit is used to minimize the path distance and iteratively optimize the cleaning path of the set of photovoltaic panels to be cleaned using a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned. The cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned. The cleaning control unit is used to control the cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path.

[0068] The working process, working details and technical effects of the photovoltaic panel adaptive automatic cleaning device provided in the second aspect of this embodiment can be found in the first aspect of the embodiment, and will not be repeated here.

[0069] Please see Figure 3 The third aspect of this application provides an electronic device, including a memory, a processor, and a transceiver that are sequentially and communicatively connected, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the photovoltaic panel adaptive automatic cleaning method as described in the first aspect of the application.

[0070] Specifically, the memory may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out (FIFO) memory, and / or last-in-first-out (FILO) memory, etc.; the processor may not be limited to microprocessors of the STM32F105 series, ARM (Advanced RISC Machines), x86 architecture processors, or processors with integrated NPU (neural-network processing units); the transceiver may be, but is not limited to, WiFi (Wireless Fidelity) wireless transceivers, Bluetooth wireless transceivers, General Packet Radio Service (GPRS) wireless transceivers, ZigBee (a low-power LAN protocol based on the IEEE 802.15.4 standard), 3G transceivers, 4G transceivers, and / or 5G transceivers, etc.

[0071] This fourth aspect of the embodiment provides a computer-readable storage medium storing instructions containing the photovoltaic panel adaptive automatic cleaning method described in the first aspect of the embodiment. Specifically, the computer-readable storage medium stores instructions that, when executed on a computer, perform the photovoltaic panel adaptive automatic cleaning method as described in the first aspect. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0072] The fifth aspect of this embodiment provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the photovoltaic panel adaptive automatic cleaning method as described in the first aspect of this embodiment, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0073] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for adaptive automatic cleaning of photovoltaic panels, characterized in that, include: Receive regional visible light images and regional thermal images of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera; The visible light images of each photovoltaic panel in the visible light image of the region and the thermal imaging images of each photovoltaic panel in the thermal imaging image of the region are extracted by the target detection algorithm. The visible light images of each photovoltaic panel are converted into grayscale images, and the grayscale histogram of the corresponding grayscale image of each photovoltaic panel is extracted. For each photovoltaic panel's grayscale histogram, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the preset grayscale range is used as the binary segmentation threshold of the grayscale image corresponding to the photovoltaic panel. Based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, the grayscale image corresponding to each photovoltaic panel is binary segmented, and the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel are taken as the pixels of the dust-covered area, thus obtaining the dust-covered area corresponding to each photovoltaic panel. Based on the thermal imaging images of each photovoltaic panel, the average temperature of the dust-covered area corresponding to each photovoltaic panel is determined. The dust coverage of the dust-covered area corresponding to each photovoltaic panel is determined based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current period. Based on the area and dust coverage of each photovoltaic panel, the overall dust coverage of each photovoltaic panel is determined. Photovoltaic panels whose overall dust coverage exceeds a preset threshold are identified as photovoltaic panels to be cleaned, thus obtaining a set of photovoltaic panels to be cleaned. With the goal of minimizing the path distance, the cleaning path of the set of photovoltaic panels to be cleaned is iteratively optimized using a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned. The cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned. The cleaning robot is controlled to clean each photovoltaic panel along the final cleaning path.

2. The adaptive automatic cleaning method for photovoltaic panels according to claim 1, characterized in that, After converting the visible light images of each photovoltaic panel into grayscale images, the method further includes: Filter the grayscale image corresponding to each photovoltaic panel; Morphological processing is performed on the filtered grayscale images corresponding to each photovoltaic panel to obtain the processed grayscale images corresponding to each photovoltaic panel. The step of extracting the grayscale histogram of the grayscale image corresponding to each photovoltaic panel includes: Extract the grayscale histogram of the processed grayscale image corresponding to each photovoltaic panel.

3. The adaptive automatic cleaning method for photovoltaic panels according to claim 1, characterized in that, For any given photovoltaic panel's grayscale histogram, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within a preset grayscale interval is used as the binary segmentation threshold for the grayscale image corresponding to the photovoltaic panel, including: For any photovoltaic panel, determine whether there are two frequency peaks in the gray-level histogram of any photovoltaic panel within the first preset gray-level interval; If the grayscale histogram corresponding to any photovoltaic panel contains two frequency peaks within the first preset grayscale interval, then the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the first preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel. If the grayscale histogram corresponding to any photovoltaic panel does not have two frequency peaks within the first preset grayscale interval, the grayscale value corresponding to the minimum frequency between the two maximum frequencies within the second preset grayscale interval is used as the binary segmentation threshold of the grayscale image corresponding to any photovoltaic panel, wherein the first preset grayscale interval is a subset of the second preset grayscale interval.

4. The adaptive automatic cleaning method for photovoltaic panels according to claim 1, characterized in that, With the objective of minimizing path distance, a genetic algorithm is used to iteratively optimize the cleaning paths for the set of photovoltaic panels to be cleaned, resulting in the final cleaning path for each panel, including: After encoding each photovoltaic panel in the set of photovoltaic panels to be cleaned, a population is initialized, and each chromosome in the population represents a cleaning path; The fitness value of each chromosome in the population is calculated based on the path distances corresponding to each chromosome in the population. The population is updated iteratively by selecting the chromosomes with the highest fitness values ​​from the population through tournament selection and selecting individuals from the remaining chromosomes of the population through roulette wheel selection for crossover and mutation operations. When the iteration stops, the cleaning path corresponding to the chromosome with the highest fitness value is taken as the final cleaning path.

5. The photovoltaic panel adaptive automatic cleaning method according to claim 4, characterized in that, The fitness value of a chromosome ,in, , , This represents the fitness value of a specific chromosome in the k-th generation. This represents the maximum path distance among all chromosomes in the k-th generation. This represents the path distance corresponding to a specific chromosome in the k-th generation. This represents the selection pressure regulation value of the chromosome in the kth generation. Let M and c represent the selection pressure regulation value of the chromosome in the (k-1)th generation. M and c are both constants, and c ∈ [0.95, 0.99].

6. The photovoltaic panel adaptive automatic cleaning method according to claim 4, characterized in that, The population is iteratively updated by selecting the chromosomes with the highest fitness values ​​from the population through tournament selection, and then selecting individuals from the remaining chromosomes through roulette wheel selection for crossover and mutation operations. This process includes: The tournament selection method is used to select the chromosomes with the highest fitness values ​​from the population as offspring chromosomes. Randomly select N chromosomes from the remaining chromosomes in the population; The offspring chromosome is obtained by selecting one chromosome from the N chromosomes using a roulette wheel selection method as the parent chromosome and performing crossover and mutation operations. N chromosomes are randomly selected from the remaining chromosomes in the population. Then, the parent chromosomes are selected from the randomly selected N chromosomes using a roulette wheel selection method. Crossover and mutation operations are performed until the number of all offspring chromosomes reaches the preset number. At this point, the population iteration update is considered complete, and all offspring chromosomes are updated to parent chromosomes for use in the next population iteration update.

7. The adaptive automatic cleaning method for photovoltaic panels according to claim 6, characterized in that, The offspring chromosomes are obtained by selecting one chromosome from the N chromosomes using a roulette wheel selection method as the parent chromosome, performing crossover and mutation operations, and including: A chromosome is selected from the N chromosomes using a roulette wheel selection method as the first parent chromosome; Determine whether the path distance corresponding to the first parent chromosome is greater than a distance threshold, wherein the distance threshold is determined based on the number of photovoltaic panels to be cleaned in the set of photovoltaic panels to be cleaned and the average distance between the photovoltaic panels to be cleaned in the set of photovoltaic panels to be cleaned; If the path distance corresponding to the first parent chromosome is greater than the distance threshold, then a chromosome is randomly selected from the N chromosomes as the second parent chromosome, and the first parent chromosome and the second parent chromosome are sequentially crossed and then mutated according to the preset mutation probability to obtain the offspring chromosome; If the path distance corresponding to the first parent chromosome is less than or equal to the distance threshold, then the first parent chromosome undergoes a crossover operation and a mutation operation is performed according to a preset mutation probability to obtain the offspring chromosome.

8. A photovoltaic panel adaptive automatic cleaning device, characterized in that, include: The receiving unit is used to receive the regional visible light image and regional thermal imaging image of the photovoltaic panel area in the photovoltaic power station acquired by the dual-light camera; The extraction unit is used to extract the visible light images of each photovoltaic panel in the visible light image of the region and the thermal imaging images of each photovoltaic panel in the thermal imaging image of the region using a target detection algorithm. The conversion unit is used to convert the visible light image of each photovoltaic panel into a grayscale image; The extraction unit is also used to extract the grayscale histogram of the grayscale image corresponding to each photovoltaic panel; The segmentation threshold selection unit is used to select the gray value corresponding to the minimum frequency between the two maximum frequencies within a preset gray range as the binary segmentation threshold of the gray-level image corresponding to the photovoltaic panel for the gray-level histogram corresponding to each photovoltaic panel. The binary segmentation unit is used to perform binary segmentation on the grayscale image corresponding to each photovoltaic panel based on the binary segmentation threshold of the grayscale image corresponding to each photovoltaic panel, and to take the pixels with a brightness value of 255 in the binary segmented image corresponding to each photovoltaic panel as the pixels of the dust-covered area, so as to obtain the dust-covered area corresponding to each photovoltaic panel. The determination unit is used to determine the average temperature of the dust-covered area corresponding to each photovoltaic panel based on the thermal imaging images of each photovoltaic panel. The determining unit is also used to determine the dust coverage of the dust-covered area corresponding to each photovoltaic panel based on the average temperature of the dust-covered area corresponding to each photovoltaic panel and the average ambient temperature during the current time period. The determining unit is also used to determine the comprehensive dust coverage of each photovoltaic panel based on the area and dust coverage of the dust-covered area corresponding to each photovoltaic panel. A photovoltaic panel selection unit is used to select photovoltaic panels whose overall dust coverage exceeds a preset threshold as photovoltaic panels to be cleaned, thereby obtaining a set of photovoltaic panels to be cleaned. The iterative optimization unit is used to minimize the path distance and iteratively optimize the cleaning path of the set of photovoltaic panels to be cleaned using a genetic algorithm to obtain the final cleaning path of each photovoltaic panel to be cleaned. The cleaning path records the cleaning order of each photovoltaic panel to be cleaned in the set of photovoltaic panels to be cleaned. The cleaning control unit is used to control the cleaning robot to clean each photovoltaic panel to be cleaned along the final cleaning path.

9. An electronic device, characterized in that, The device includes a memory, a processor, and a transceiver that are sequentially and communicatively connected. The memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the photovoltaic panel adaptive automatic cleaning method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or the instructions are executed by the computer, they implement the photovoltaic panel adaptive automatic cleaning method as described in any one of claims 1 to 7.

Images