Photovoltaic panel intelligent cleaning method and system for photovoltaic operation and maintenance

Abstract

The invention relates to the field of photovoltaic panel cleaning, and discloses a photovoltaic panel intelligent cleaning method and system for photovoltaic operation and maintenance, and the method comprises the steps: carrying out the detection of the state of a photovoltaic panel in a preset time period, detecting the power generation amount, recognizing the dirt distribution, hot spot defects or physical damage, and recording the position of an abnormal region, and obtaining a detection result; a multi-parameter cleaning decision model is established, the detection result is imported into the decision model, and whether a cleaning system is started or not is judged; when it is judged that the cleaning system is started, the photovoltaic panel is cleaned according to the detection result; after cleaning is completed, the state of the photovoltaic panel is subjected to secondary detection through the detection equipment, the generating capacity is detected, and it is confirmed that no stain residues or newly-added damage exists; and after the secondary detection is completed, a detection result and a cleaning result are input into the system, and a data report is generated and uploaded to an upper computer, so that the water, electricity and labor costs are saved, and the precision, automation and intelligentization of the cleaning operation and maintenance of the photovoltaic power station are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of photovoltaic panel cleaning, in particular to a photovoltaic panel intelligent cleaning method and system for photovoltaic operation and maintenance. BACKGROUND

[0002] With the rapid development of photovoltaic power generation technology, the scale of photovoltaic power stations is continuously expanding, and the cleaning and maintenance of photovoltaic panels has become increasingly prominent. The accumulation of dust, snow, bird droppings and other pollutants on the surface of photovoltaic panels can significantly reduce their photoelectric conversion efficiency, resulting in a loss of power generation. Studies have shown that the power generation efficiency of severely polluted photovoltaic panels can be reduced by more than 20%, and even cause local hot spot effects, shortening the service life of the components.

[0003] However, the present application inventors found that the above-mentioned technology at least has the following technical problems in the process of implementing the technical solutions of the embodiments of the present application:

[0004] Traditional photovoltaic panel cleaning methods mainly rely on manual cleaning or fixed spraying systems, which have low efficiency, high cost, and waste of water resources. Manual cleaning requires frequent scheduling of personnel, especially in large photovoltaic power stations or distributed rooftop photovoltaic scenarios, making it difficult to achieve full coverage. Fixed spraying systems are limited by water supply and regional climate conditions, and may form water stains on the panel surface due to water quality problems. In recent years, some companies have attempted to use cleaning robots or drones for automated cleaning, but still face challenges such as inaccurate path planning, unstable cleaning effect, and insufficient equipment reliability. SUMMARY

[0005] The embodiments of the present application provide a photovoltaic panel intelligent cleaning method and system for photovoltaic operation and maintenance, which solves the technical problems of low intelligence level, low cleaning efficiency, single decision basis and lack of effect verification mechanism of traditional cleaning methods for photovoltaic panels in the prior art, and improves the power generation efficiency, reduces the operation and maintenance cost, and prolongs the service life of the equipment.

[0006] In order to achieve the above-mentioned purpose, the present application provides a photovoltaic panel intelligent cleaning method for photovoltaic operation and maintenance, comprising:

[0007] detecting the state of the photovoltaic panel in a preset time period, detecting the power generation, identifying the dirt distribution, hot spot defects or physical damage, and recording the position of the abnormal area to obtain a detection result;

[0008] establishing a multi-parameter cleaning decision model, importing the detection result into the decision model, and determining whether to start the cleaning system;

[0009] when it is determined to start the cleaning system, cleaning the photovoltaic panel according to the detection result;

[0010] After the cleaning is completed, the state of the photovoltaic panel is detected again by the detection device, the power generation is detected, and it is confirmed that there is no dirt residue and no new damage;

[0011] After the secondary detection is completed, the detection results and the cleaning results are entered into the system, and a data report is generated and uploaded to the upper computer.

[0012] Further, when detecting the state of the photovoltaic panel by the detection device and the preset time period, it includes:

[0013] The infrared thermal imager is used to identify the temperature abnormal area on the surface of the photovoltaic panel to locate the hot spot defect;

[0014] The high-definition camera device is used to shoot the image of the surface of the photovoltaic panel, and the image recognition algorithm is used to analyze the type, thickness and distribution area of the dirt;

[0015] The power generation efficiency data of the photovoltaic panel is monitored and recorded in real time by the electrical performance testing device;

[0016] The coordinate position of the identified hot spot, dirt concentration area or physical damage is marked and stored by the positioning module.

[0017] Further, when the high-definition camera device is used to shoot the image of the surface of the photovoltaic panel, and the image recognition algorithm is used to analyze the type, thickness and distribution area of the dirt, it includes:

[0018] A target detection model based on deep learning is used to segment and identify the collected image to distinguish different types of dirt such as dust, bird droppings and oil stains;

[0019] A dirt coverage thickness estimation model is established by analyzing the pixel gray value, texture features and contrast of the dirt area in the image and the background;

[0020] The pixel coordinates and actual positions of the identified dirt area are mapped and calculated in combination with the actual size data of the photovoltaic panel to determine the distribution range and area ratio of the dirt area.

[0021] Further, a multi-parameter cleaning decision model is established, and the detection results are input into the decision model to determine whether to start the cleaning system, including:

[0022] A multi-parameter cleaning decision model is established, and three core parameters of power generation efficiency δ, dirt coverage rate α and hot spot area ratio β in the detection results are comprehensively judged;

[0023] A first threshold condition is set: when the power generation efficiency δ is less than or equal to 95%, it is determined that cleaning is needed;

[0024] A second threshold condition is set: when the dirt coverage rate α is greater than or equal to 15%, it is determined that cleaning is needed;

[0025] A third threshold condition is set: when the hot spot area ratio β is greater than or equal to 2%, it is determined that cleaning is needed;

[0026] A weighted decision function F = w1 x (1-δ) + w2 x α + w3 x β is used for comprehensive evaluation, wherein the weight coefficients w1 = 0.5, w2 = 0.3, and w3 = 0.2;

[0027] When the weighted decision function F is greater than or equal to 4, the cleaning system is automatically started.

[0028] Further, regarding the determination method of the power generation efficiency δ, the method comprises:

[0029] The power generation amount corresponding to each collection time period and the theoretical power generation amount are determined;

[0030] A random power generation amount is extracted, and all left and right power generation amounts corresponding to the power generation amount are extracted, wherein when the power generation amount is randomly extracted, the power generation amount corresponding to the initial time and the power generation amount corresponding to the end time are not included;

[0031] The first power generation mean of all left power generation amounts is calculated, and the second power generation mean of all right power generation amounts is calculated;

[0032] According to the relationship among the power generation amount, the first power generation mean, and the second power generation mean, the corresponding power generation change coefficient is calculated;

[0033] All power generation change coefficients are extracted, and the mean of all power generation change coefficients is taken as the power generation efficiency δ, and the cleaning strategy of the photovoltaic panel is set based on the power generation efficiency δ;

[0034] The power generation change coefficient is calculated according to the following formula:

[0035]

[0036] Wherein n is the power generation change coefficient, m is the power generation amount, the minimum value of the first power generation mean and the second power generation mean is taken as m2, and the maximum value of the first power generation mean and the second power generation mean is taken as m1.

[0037] Further, the cleaning strategy of the photovoltaic panel is set based on the power generation efficiency δ, which comprises:

[0038] A first preset power generation efficiency and a second preset power generation efficiency are preset;

[0039] A first preset optimization value, a second preset optimization value, and a third preset optimization value are preset;

[0040] When the power generation efficiency is greater than the first preset power generation efficiency, the first preset optimization value and the first product value of the cleaning flow and the cleaning pressure are calculated respectively to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel.

[0041] When the power generation efficiency is less than or equal to the first preset power generation efficiency and greater than the second preset power generation efficiency, the second preset optimization value and the second product value of the cleaning flow and the cleaning pressure are calculated respectively to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel.

[0042] When the power generation efficiency is less than or equal to the second preset power generation efficiency, the third preset optimization value and the third product value of the cleaning flow and the cleaning pressure are calculated respectively to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel.

[0043] Further, the photovoltaic panel is cleaned according to the detection result, comprising:

[0044] The photovoltaic panel is pre-cleaned using a soft brush or compressed air to handle loose dirt on the surface, including dust, leaves and particulate matter;

[0045] The dirt concentration area or hot spot area is cleaned in a targeted manner;

[0046] The surface of the photovoltaic panel is cleaned as a whole.

[0047] Further, the dirt concentration area or hot spot area is cleaned in a targeted manner, specifically comprising:

[0048] According to the detection result, the precise distribution position of the dirt concentration area or hot spot area is determined by a pre-trained dirt coverage thickness estimation model, and the dirt adhesion grade d is calculated;

[0049] Based on the dirt adhesion grade d, the cleaning water flow speed v is dynamically adjusted according to the tool speed control formula v = f(d), wherein the function f(d) is a nonlinear positive correlation function based on the basic speed and the dirt thickness;

[0050] According to the adjusted speed parameter, the dirt concentration area or hot spot area is cleaned in a targeted and accurate manner.

[0051] Further, after the secondary detection is completed, the detection result and the cleaning result are input into the system, and a data report is generated and uploaded to the upper computer, wherein the data report comprises:

[0052] The decision parameters δ, α, β, F and the decision result are stored in association with the time stamp, and a cleaning decision report is generated;

[0053] Record the cleaning tool type, working pressure / rotation speed, moving speed, cleaning times and cleaning effect evaluation data, and generate a cleaning process report.

[0054] To achieve the above object, the application also provides a temperature control system for a vertical furnace, further comprising:

[0055] A data detection module is configured to detect the state of the photovoltaic panel in a preset time period, detect the power generation, identify the dirt distribution, hot spot defects or physical damage, and record the position of the abnormal area to obtain a detection result.

[0056] A data judgment module is configured to establish a multi-parameter cleaning decision model, import the detection result into the decision model, and judge whether to start the cleaning system.

[0057] A cleaning execution module is configured to clean the photovoltaic panel according to the detection result when it is judged to start the cleaning system.

[0058] A secondary detection module is configured to perform secondary detection on the state of the photovoltaic panel by the detection device after cleaning, detect the power generation, and confirm that there is no dirt residue and no new damage.

[0059] A data processing module is configured to record the detection result and the cleaning result into the system and generate a data report after the secondary detection is completed, and upload the data report to an upper computer.

[0060] Compared with the prior art, the application has the following beneficial effects:

[0061] The state of the photovoltaic panel is detected in a preset time period, the power generation is detected, the dirt distribution, hot spot defects or physical damage are identified, and the position of the abnormal area is recorded to obtain a detection result. BRIEF DESCRIPTION OF DRAWINGS

[0062] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Moreover, the same reference numerals in the attached drawings refer to the same or similar components throughout the several drawings. In the drawings:

[0063] Figure 1A flowchart of a photovoltaic panel intelligent cleaning method for photovoltaic operation and maintenance in an embodiment of the application is shown.

[0064] Figure 2 A structural diagram of a photovoltaic panel intelligent cleaning system for photovoltaic operation and maintenance in an embodiment of the application is shown. DETAILED DESCRIPTION

[0065] The specific embodiments of the application will be further described in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the application, but are not used to limit the scope of the application.

[0066] In the description of the present application, it should be understood that the terms 'center', 'upper', 'lower', 'front', 'back', 'left', 'right','vertical', 'horizontal', 'top', 'bottom', 'inner', 'outer' and the like indicate the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0067] The terms 'first','second' are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with 'first','second' can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of 'a plurality of' is two or more.

[0068] In the description of the present application, it should be noted that, unless otherwise specified and limited, the terms'mounting', 'connecting', 'connection' should be understood broadly, for example, it can be fixed connection, or detachable connection, or integral connection; it can be mechanical connection, or electrical connection; it can be direct connection, or indirect connection through intermediate medium, or internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0069] The following is a description of the preferred embodiments of the application in conjunction with the accompanying drawings.

[0070] As Figure 1 shown, the embodiments of the application disclose a photovoltaic panel intelligent cleaning method for photovoltaic operation and maintenance, comprising:

[0071] S110: detecting the state of the photovoltaic panel in a preset time period, detecting the power generation, identifying the dirt distribution, thermal spot defects or physical damage, and recording the position of the abnormal area to obtain the detection result.

[0072] In this embodiment, the preset time period is set to a period with low and stable irradiance, such as early morning or overcast day, during which the surface temperature of the photovoltaic panel is uniform and low, which can greatly improve the recognition accuracy of the infrared thermal imager for hot spot defects and avoid the interference of high temperature background noise under strong light; at the same time, the soft diffuse light environment is conducive to the high-definition camera to obtain clear surface images without shadows and reflections, thereby ensuring the accuracy of the image recognition algorithm for analyzing the type, thickness and distribution of dirt; in addition, the power generation data of this period is least affected by sunshine fluctuations, which can provide the most reliable reference parameters for calculating the power generation efficiency reduction ratio.

[0073] In this embodiment, when detecting the power generation, identifying the dirt distribution, hot spot defects or physical damage, and recording the positions of abnormal areas, the following steps are further included:

[0074] identifying temperature abnormal areas on the surface of the photovoltaic panel using an infrared thermal imager to locate hot spot defects;

[0075] capturing images of the surface of the photovoltaic panel by a high-definition camera and analyzing the type, thickness and distribution of dirt using an image recognition algorithm;

[0076] real-time monitoring and recording the power generation efficiency data of the photovoltaic panel using an electrical performance testing device;

[0077] marking and storing the coordinate positions of the identified hot spots, dirt concentration areas or physical damage by a positioning module.

[0078] In this embodiment, the training process of the image recognition algorithm is as follows:

[0079] Data preparation: Use a high-definition camera to capture a large number of high-definition images of the surface of the photovoltaic panel under different seasons, different weather conditions and different light conditions. It is necessary to ensure that various types of dirt, different degrees of dirt thickness, hot spots and physical damage are covered.

[0080] Data cleaning: eliminate invalid images that are blurred, too dark, overexposed or have serious obstructions.

[0081] Feature extraction: label the dirt distribution, thickness and type.

[0082] Model construction: select a YOLO target detection model to realize fast positioning and classification of dirt.

[0083] Model training: divide the prepared data into training set, validation set and test set, and put the training set into the data model for training to improve the judgment accuracy of the model; use the validation set to monitor the training process to prevent overfitting, and adjust the learning rate, batch size and other hyperparameters to select the best model on the validation set; finally, test the test set to ensure the accuracy of the model before putting it into use.

[0084] The beneficial effects of the above technical solution are: by intelligently presetting the detection period to the morning or overcast day and the like, the irradiance is low and stable, which significantly improves the accuracy and reliability of state detection. Under this period, the uniform low-temperature surface of the photovoltaic panel enables the infrared thermal imager to clearly identify hot spot defects and avoid strong light interference; the soft diffuse light environment ensures that the high-definition camera obtains clear images without glare and shadow, providing ideal conditions for accurate analysis of dirt type, thickness and distribution based on deep learning image recognition algorithms; at the same time, stable power generation data provides a reliable benchmark for calculating efficiency reduction. Combined with multi-sensor fusion detection (infrared, visual, electrical performance) and precise positioning technology, the present scheme realizes comprehensive quantitative perception and accurate positioning of photovoltaic panel dirt, hot spots, damage and power generation performance, providing an efficient and accurate data basis for subsequent cleaning decisions, and fundamentally solving the pain points of traditional detection methods being greatly disturbed by the environment and not being accurate in recognition, greatly improving the intelligent level and decision-making scientificity of photovoltaic operation and maintenance.

[0085] S120: Establish a multi-parameter cleaning decision-making model, import the detection results into the decision-making model, and determine whether to start the cleaning system.

[0086] In some embodiments of the present application, before establishing a multi-parameter cleaning decision-making model, importing the detection results into the decision-making model, and determining whether to start the cleaning system, the method further comprises:

[0087] Establishing a multi-parameter cleaning decision-making model, and comprehensively judging based on three core parameters of power generation efficiency δ, dirt coverage rate α and hot spot area ratio β in the detection results;

[0088] Setting a first threshold condition: when the power generation efficiency δ is less than or equal to 95%, it is determined that cleaning is needed;

[0089] Setting a second threshold condition: when the dirt coverage rate α is greater than or equal to 15%, it is determined that cleaning is needed;

[0090] Setting a third threshold condition: when the hot spot area ratio β is greater than or equal to 2%, it is determined that cleaning is needed;

[0091] Comprehensively evaluating by using a weighted decision function F = w1 x (1- δ) + w2 x α + w3 x β, wherein the weight coefficients w1 = 0.5, w2 = 0.3, and w3 = 0.2;

[0092] When the weighted decision function F is greater than or equal to 4, the cleaning system is automatically started.

[0093] In the present embodiment, w1 = 0.5, w2 = 0.3, and w3 = 0.2 are preferred, and can be adaptively adjusted according to actual conditions.

[0094] In some embodiments of the present application, the determination of the power generation efficiency δ is as follows, comprising:

[0095] determining the power generation of each collection time period and the theoretical power generation;

[0096] randomly extracting a power generation, extracting all left and right power generations corresponding to the power generation, wherein when the power generation is randomly extracted, the power generation corresponding to the initial time and the power generation corresponding to the end time are not included;

[0097] calculating the first power generation average of all left power generations and the second power generation average of all right power generations;

[0098] calculating the corresponding power generation change coefficient according to the relationship among the power generation, the first power generation average and the second power generation average;

[0099] extracting all power generation change coefficients, and taking the average of all power generation change coefficients as the power generation efficiency δ, and setting the cleaning strategy of the photovoltaic panel based on the power generation efficiency δ;

[0100] calculating the power generation change coefficient according to the following formula:

[0101]

[0102] wherein n is the power generation change coefficient, m is the power generation, the minimum of the first power generation average and the second power generation average is taken as m2, and the maximum of the first power generation average and the second power generation average is taken as m1.

[0103] In the present embodiment, the values of m1 and m2 adopt a dynamic reference strategy: m1 takes the larger one of the left average and the right average, representing the upper limit of the normal power generation level expected to be reached in the surrounding; m2 takes the smaller one of the two, representing the lower limit of the expected power generation level in the surrounding.

[0104] In some embodiments of the present application, setting the cleaning strategy of the photovoltaic panel based on the power generation efficiency δ, comprising:

[0105] pre-setting a first preset power generation efficiency and a second preset power generation efficiency;

[0106] pre-setting a first preset optimization value, a second preset optimization value and a third preset optimization value;

[0107] when the power generation efficiency is greater than the first preset power generation efficiency, then the first preset optimization value and the first product value of the cleaning flow and the cleaning pressure are calculated respectively, to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel;

[0108] When the power generation efficiency is less than or equal to the first preset power generation efficiency and greater than the second preset power generation efficiency, the second preset optimization value and the second product value of the cleaning flow and the cleaning pressure are calculated respectively to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel.

[0109] When the power generation efficiency is less than or equal to the second preset power generation efficiency, the third preset optimization value and the third product value of the cleaning flow and the cleaning pressure are calculated respectively to obtain the optimized cleaning flow and the optimized cleaning pressure as the cleaning strategy of the photovoltaic panel.

[0110] In the embodiment, the first preset power generation efficiency is preferably 0.85, and the second preset power generation efficiency is preferably 0.65, which can also be adaptively adjusted according to actual conditions.

[0111] In the embodiment, the first preset optimization value is preferably 1.1, the second preset optimization value is preferably 1.4, and the third preset optimization value is 1.8, which can also be adaptively adjusted according to actual conditions.

[0112] The beneficial effects of the above technical solution are: by comparing the real-time power generation m with the dynamic fluctuation interval composed of m1 and m2, the normal fluctuation caused by environmental factors and the abnormal attenuation caused by dirt coverage can be effectively distinguished, so that the real performance change coefficient of the power generation is accurately extracted, providing an anti-interference quantitative basis for calculating the comprehensive power generation efficiency δ, and the preset power generation efficiency level is used to intelligently associate the power generation efficiency quantitative index with the cleaning intensity parameter, so that the accurate adaptive regulation of the cleaning strategy is realized: the higher the power generation efficiency, the more gentle the cleaning mode to save resources, and the lower the power generation efficiency, the stronger the cleaning parameters are automatically triggered to ensure thorough cleaning, which effectively solves the pain points of "excessive cleaning" or "insufficient cleaning" in the traditional fixed parameter cleaning mode.

[0113] S130: When it is judged to start the cleaning system, the photovoltaic panel is cleaned according to the detection result.

[0114] In some embodiments of the present application, when it is judged to start the cleaning system, the photovoltaic panel is cleaned according to the detection result, including:

[0115] The photovoltaic panel is pre-cleaned using a soft brush or compressed air to handle loose dirt on the surface, including dust, leaves and particulate matter;

[0116] The dirt concentration area or hot spot area is cleaned in a targeted manner;

[0117] The surface of the photovoltaic panel is cleaned as a whole.

[0118] In this embodiment, the pre-cleaning uses a soft-bristle roller to make light physical contact with the photovoltaic panel surface, or uses compressed air to generate an air flow for preliminary cleaning. This method can efficiently remove light debris such as dust, fallen leaves, pollen, and sand, and the soft material and non-high-pressure method ensure that the photovoltaic glass surface will not be scratched or damaged, creating a clean environment for subsequent deep cleaning.

[0119] In this embodiment, targeted cleaning uses higher pressure and closer distance to spray water or special cleaning agents on the target area based on the detection results.

[0120] In this embodiment, the overall cleaning uses a mode that covers the entire panel surface and a standardized pressure flow to clean the photovoltaic panel comprehensively. This not only removes possible minor stains that may be missed, but also ensures the uniformity of the panel cleaning degree, avoiding the formation of new shadows or hot spots due to residual water stains or dirt, and ultimately ensuring the uniform recovery of the overall power generation performance of the photovoltaic panel.

[0121] The beneficial effects of the above technical solution are: through the step-by-step cleaning mode, precise and efficient, resource-saving, and equipment protection are achieved. At the same time, the power generation efficiency is significantly improved, water consumption, energy consumption, and equipment wear and tear are minimized, and the work efficiency is improved.

[0122] S140: After cleaning, the detection device detects the state of the photovoltaic panel again, detects the power generation, and confirms that there is no stain residue and new damage.

[0123] In this embodiment, after cleaning, the system calls the infrared thermal imager, high-definition camera device, and electrical performance test equipment again to perform full-dimensional re-inspection on the photovoltaic panel. By comparing the power generation data before and after cleaning to verify the efficiency recovery degree, using image recognition algorithms to scan the surface to confirm that there is no stain residue, and using infrared detection to check for new physical damage or hot spot hazards that may be caused during the cleaning process.

[0124] The beneficial effects of the above technical solution are: through the secondary detection after cleaning, the verifiable recovery of the power generation performance is ensured, and a digital cleaning file is formed to provide data support for operation and maintenance optimization, ultimately realizing significant value in extending the service life of the components and ensuring power generation revenue.

[0125] S150: After the secondary detection is completed, the detection results and cleaning results are entered into the system, and a data report is generated and uploaded to the upper computer.

[0126] In some embodiments of the present application, after the secondary detection is completed, the detection results and cleaning results are entered into the system, and a data report is generated and uploaded to the upper computer, wherein the data report includes:

[0127] The decision parameters delta, alpha, beta, F and the decision result are stored in association with the time stamp, and a cleaning decision report is generated;

[0128] The cleaning tool type, working pressure / rotation speed, moving speed, cleaning frequency and cleaning effect evaluation data are recorded, and a cleaning process report is generated.

[0129] In this embodiment, the cleaning decision report records the trigger logic (parameters such as delta, alpha, beta, F and decision result) of this task, and is bound with the time stamp to form a traceable decision archive.

[0130] In this embodiment, the cleaning process report records the type, pressure, speed, frequency and final effect evaluation of the cleaning equipment, and the execution data.

[0131] The beneficial effects of the above technical solution are: the entire operation and cleaning process is digitized, providing a data basis for subsequent fault prediction and operation strategy of the photovoltaic panel, and providing data support for subsequent operation iteration.

[0132] In order to further illustrate the technical idea of the present application, the technical solution of the present application will be described in conjunction with a specific application scenario.

[0133] Correspondingly, as shown in Figure 2 The present application also provides a photovoltaic panel intelligent cleaning system for photovoltaic operation, comprising:

[0134] A data detection module is configured to detect the state of the photovoltaic panel in a preset time period, detect the power generation, identify the dirt distribution, thermal spot defects or physical damage, and record the position of the abnormal area to obtain a detection result.

[0135] A data judgment module is configured to establish a multi-parameter cleaning decision model, import the detection result into the decision model, and judge whether to start the cleaning system.

[0136] A cleaning execution module is configured to clean the photovoltaic panel according to the detection result when it is judged to start the cleaning system.

[0137] A secondary detection module is configured to perform secondary detection on the state of the photovoltaic panel through the detection device after cleaning, detect the power generation, and confirm that there is no dirt residue and no new damage.

[0138] A data processing module is configured to record the detection result and cleaning result into the system after the secondary detection is completed, and generate a data report and upload it to the upper computer.

[0139] In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0140] While the application has been described above with reference to the examples, it is apparent that various modifications and changes can be made to the application without departing from the scope thereof. In particular, any feature described in relation to one embodiment can be used in combination with any other feature described in relation to any other embodiment, unless the structures conflict. The scope of the application is not limited to the specific examples described herein, but only by the claims.

[0141] It is apparent that the above only describes preferred embodiments of the application and not limit the application. Although the application has been described in detail with reference to the foregoing embodiments, the technical solutions recorded in the foregoing embodiments can still be modified or equivalent replaced by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A method for intelligent cleaning of photovoltaic panels for photovoltaic operation and maintenance, characterized in that, include: The photovoltaic panel status is detected within a preset time period to detect power generation, identify dirt distribution, hot spot defects or physical damage, and record the location of abnormal areas to obtain the detection results. A multi-parameter cleaning decision model is established, and the detection results are imported into the decision model to determine whether to start the cleaning system. When it is determined that the cleaning system should be activated, the photovoltaic panels are cleaned according to the detection results; After cleaning, the condition of the photovoltaic panel is tested a second time using the testing equipment to check the power generation and confirm that there are no stains or new damage. After the second test is completed, the test results and cleaning results are entered into the system, a data report is generated, and uploaded to the host computer.

2. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 1, characterized in that, When detecting the state of the photovoltaic panel using detection equipment and within a preset time period, the following are included: An infrared thermal imager is used to identify temperature anomaly areas on the surface of the photovoltaic panel in order to locate hot spot defects. The photovoltaic panel surface is captured by a high-definition camera, and the type, thickness, and distribution area of ​​dirt are analyzed using image recognition algorithms. The power generation efficiency data of the photovoltaic panel is monitored and recorded in real time using an electrical performance testing device. The positioning module marks and stores the coordinates of identified hot spots, areas of concentrated dirt, or physical damage.

3. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 2, characterized in that, When capturing images of the photovoltaic panel surface using a high-definition camera and analyzing the type, thickness, and distribution area of ​​dirt using image recognition algorithms, the process includes: A deep learning-based object detection model is used to segment and identify the acquired images to distinguish different types of dirt such as dust, bird droppings, and oil stains. By analyzing the pixel grayscale values, texture features, and contrast between the image and the background of the dirt area in the image, a model for estimating the dirt coverage thickness is established. Based on the actual size data of the photovoltaic panel, the pixel coordinates of the identified dirt areas are mapped to their actual locations to determine the distribution range and area ratio of the dirt areas.

4. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 1, characterized in that, Establish a multi-parameter cleaning decision model, import the detection results into the decision model, and determine whether to start the cleaning system, including: A multi-parameter clean decision model is established, and a comprehensive judgment is made based on three core parameters in the test results: power generation efficiency δ, dirt coverage α, and hot spot area ratio β. Set a first threshold condition: when the power generation efficiency δ≤95%, it is determined that cleaning is required; Set a second threshold condition: when the dirt coverage α ≥ 15%, it is determined that cleaning is required; A third threshold condition is set: when the hot spot area ratio β ≥ 2%, it is determined that cleaning is required; A weighted decision function F = w1 × (1 - δ) + w2 × α + w3 × β is used for comprehensive evaluation, where the weight coefficients w1 = 0.5, w2 = 0.3, and w3 = 0.

2. When the weighted decision function F is greater than or equal to 4, the cleaning system is automatically started.

5. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 4, characterized in that, The power generation efficiency δ is determined as follows, including: Determine the power generation and theoretical power generation for each data collection period; Randomly extract a power generation value, and extract all left-side power generation values ​​and right-side power generation values ​​corresponding to the power generation value. When randomly extracting power generation values, the power generation values ​​corresponding to the initial time and the power generation values ​​corresponding to the end time are not included. Calculate the first average of all power generation on the left side, and calculate the second average of all power generation on the right side. Based on the relationship between the power generation, the first average power generation, and the second average power generation, calculate the corresponding power generation variation coefficient; Extract all power generation variation coefficients and take the average value corresponding to all power generation variation coefficients as the power generation efficiency δ. Set the cleaning strategy for the photovoltaic panel based on the power generation efficiency δ. The coefficient for change in power generation is calculated using the following formula: Where n is the power generation variation coefficient, m is the power generation, the minimum of the first power generation average and the second power generation average is taken as m2, and the maximum of the first power generation average and the second power generation average is taken as m1.

6. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 5, characterized in that, The cleaning strategy for the photovoltaic panel is set based on the power generation efficiency δ, including: The first preset power generation efficiency and the second preset power generation efficiency are preset. Pre-set the first preset optimization value, the second preset optimization value, and the third preset optimization value; When the power generation efficiency is greater than the first preset power generation efficiency, the first preset optimized value and the first product value of the cleaning flow rate and cleaning pressure are calculated respectively to obtain the optimized cleaning flow rate and optimized cleaning pressure, which are used as the cleaning strategy for the photovoltaic panel. When the power generation efficiency is less than or equal to the first preset power generation efficiency and greater than the second preset power generation efficiency, the second preset optimized value and the second product value of the cleaning flow rate and cleaning pressure are calculated respectively to obtain the optimized cleaning flow rate and optimized cleaning pressure, which are used as the cleaning strategy for the photovoltaic panel. When the power generation efficiency is less than or equal to the second preset power generation efficiency, the third preset optimized value and the third product value of the cleaning flow rate and cleaning pressure are calculated respectively to obtain the optimized cleaning flow rate and optimized cleaning pressure, which are used as the cleaning strategy for the photovoltaic panel.

7. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 1, characterized in that, The photovoltaic panel is cleaned according to the test results, including: The photovoltaic panels are pre-cleaned using a soft brush or compressed air to remove loose dirt, including dust, leaves and particles, from the surface. Targeted cleaning of areas with concentrated dirt or hot spots; The surface of the photovoltaic panel is thoroughly cleaned.

8. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 7, characterized in that, Targeted cleaning of areas with concentrated dirt or hot spots, specifically including: Based on the detection results, the precise distribution location of the dirt concentration area or hot spot area is determined by a pre-trained dirt coverage thickness estimation model, and its dirt adhesion level d is calculated. Based on the aforementioned dirt adhesion level d, the cleaning water flow speed v is dynamically adjusted according to the tool speed control formula v = f(d), where the function f(d) is a nonlinear positive correlation function based on the base speed and dirt thickness. Based on the adjusted speed parameters, the areas with concentrated dirt or hot spots are precisely cleaned in a targeted manner.

9. The intelligent cleaning method for photovoltaic panels for photovoltaic operation and maintenance according to claim 1, characterized in that, After the secondary inspection is completed, the inspection results and cleaning results are entered into the system, a data report is generated, and uploaded to the host computer. The data report includes: The decision parameters δ, α, β, F and the decision results are associated with and stored with timestamps, and a cleaning decision report is generated. Record cleaning tool type, working pressure / speed, moving speed, number of cleaning cycles, and cleaning effect evaluation data to generate a cleaning process report.

10. A photovoltaic panel intelligent cleaning system for photovoltaic operation and maintenance, applied to the photovoltaic panel intelligent cleaning method for photovoltaic operation and maintenance as described in any one of claims 1-9, characterized in that, include: The data detection module is used to detect the status of the photovoltaic panel within a preset time period, detect power generation, identify dirt distribution, hot spot defects or physical damage, record the location of abnormal areas, and obtain the detection results. The data judgment module is used to establish a multi-parameter cleaning decision model, import the detection results into the decision model, and determine whether to start the cleaning system. The cleaning execution module is used to clean the photovoltaic panel according to the detection results when it is determined that the cleaning system will be started. The secondary inspection module is used to perform a secondary inspection of the photovoltaic panel's condition after cleaning, using the inspection equipment to detect the power generation and confirm that there are no stains or new damage. The data processing module is used to input the test results and cleaning results into the system after the secondary test is completed, generate a data report, and upload it to the host computer.

Images