Photovoltaic panel cleanliness detection method

Abstract

The invention relates to the technical field of photovoltaic panel cleanliness detection, in particular to a photovoltaic panel cleanliness detection method which comprises the steps of data acquisition, image preprocessing, pollution feature extraction, cleanliness index calculation, cleanliness judgment, data output and result recording and feedback. According to the invention, pollution data is acquired through imaging equipment, and adaptive filtering and morphological analysis are combined, so that precise pollution feature extraction, detection precision improvement and system real-time calculation of a cleanliness index are realized, pollution conditions are fed back through a cloud platform or an Internet of Things interface, and the power generation efficiency of a photovoltaic panel is ensured. The system can also automatically trigger a cleaning device and control the cleaning process according to a cleanliness threshold value, full-automatic management is achieved, the labor burden is relieved, in addition, a machine learning algorithm automatically adjusts the threshold value according to environment monitoring data and seasonal changes, and parallel detection further optimizes the efficiency, shortens the detection time, improves the operation and maintenance effect and prolongs the service life of the photovoltaic panel.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic panel cleanliness detection, and in particular to a photovoltaic panel cleanliness detection method. Background Art

[0002] In photovoltaic power generation systems, the cleanliness of photovoltaic panels has a crucial impact on their power generation efficiency. With the widespread application of photovoltaic power generation technology, pollutants such as dust, bird droppings, and leaves are easily accumulated on the surface of photovoltaic panels. These pollutants will cause the photovoltaic panels' photoelectric conversion efficiency to decrease, thereby affecting the power generation capacity of the entire system. Traditional photovoltaic panel cleaning methods usually rely on manual inspection or regular cleaning, which is not only time-consuming and labor-intensive, but also unable to promptly and effectively detect and deal with pollution problems, resulting in photovoltaic panels being in an inefficient working state for a long time, increasing maintenance costs.

[0003] With the continuous expansion of photovoltaic power generation, automated and intelligent photovoltaic panel cleanliness detection and maintenance methods have gradually become an urgent need for the development of the industry. Existing cleanliness detection technologies mostly rely on manual visual inspection or analysis of single spectral images, which is difficult to fully and accurately reflect the pollution of photovoltaic panels, especially in complex environments and large-scale photovoltaic fields. The limitations of traditional methods are particularly prominent. Therefore, there is an urgent need for an efficient, accurate and automated photovoltaic panel cleanliness detection technology to improve the timeliness and reliability of detection and provide an effective basis for triggering the automatic cleaning system. Summary of the invention

[0004] The purpose of the present invention is to solve the technical problems raised in the above background technology.

[0005] The present invention adopts the following technical solution: a photovoltaic panel cleanliness detection method, comprising the following steps: S1: Data acquisition, obtaining image data of the photovoltaic panel surface through imaging equipment installed around or above the photovoltaic panel, so as to comprehensively collect the pollution conditions on the photovoltaic panel surface; S2: Image preprocessing, preprocessing the image data, specifically including: Image denoising: remove random noise and light interference from images; Image grayscale transformation: convert image data from color space to grayscale image to facilitate subsequent feature extraction operations; Edge enhancement: Use the Sobel operator or Canny edge detection algorithm to perform edge enhancement on the grayscale image to highlight the contamination boundary features on the surface of the photovoltaic panel; S3: Pollution feature extraction, extracting pollution features on the surface of the photovoltaic panel from the preprocessed image, the pollution features include grayscale changes, light reflection unevenness, surface spot distribution, dust coverage area, etc., and feature extraction is achieved through morphological operation image analysis method; S4: cleanliness index calculation, based on the extracted pollution features, the cleanliness index of the photovoltaic panel is calculated, specifically, the various pollution features are weighted and summed through a preset weight model, and the result is normalized to the range of 0 to 1, where 0 represents complete cleanliness and 1 represents severe pollution; S5: Cleanliness judgment: compare the calculated cleanliness index with the preset cleanliness threshold. If the cleanliness index is higher than the threshold, it is judged that the photovoltaic panel needs to be cleaned and a cleaning prompt signal is output. If the cleanliness index is lower than the threshold, it is judged that no cleaning is required. S6: Data output: output the cleanliness index and cleaning judgment results to the user or automatic control system through the display screen, cloud platform or IoT interface to facilitate timely decision-making and cleaning operations; S7: Result recording and feedback: The test results are recorded in the database for subsequent analysis and cleaning plan formulation, and the parameter configuration of the imaging equipment is optimized through the feedback mechanism to improve the accuracy and efficiency of the test.

[0006] Preferably, the imaging device includes at least one of a visible light camera, an infrared camera, an ultraviolet camera, and a multi-spectral sensor, for acquiring surface contamination conditions of photovoltaic panels within different spectral ranges.

[0007] Preferably, the image denoising step adopts an adaptive median filter or a Gaussian filter algorithm, and dynamically adjusts the filter parameters according to the noise characteristics of the environment where the photovoltaic panel is located to ensure image quality.

[0008] Preferably, the pollution feature extraction step is performed by a morphological analysis method, including applying morphological operations such as dilation, erosion, opening and closing operations to the grayscale image to extract the structural features and distribution features of the pollution area.

[0009] Preferably, the calculation of the cleanliness index is implemented through a weighted model, which includes a weighted combination of parameters such as grayscale change, surface brightness difference, spot distribution area, surface reflectivity, etc., and the weight parameters are adjusted according to the environmental characteristics of different regions.

[0010] Preferably, the setting of the cleanliness threshold is dynamically adjusted according to historical cleaning data and environmental monitoring data, and the threshold is set in a manner including a machine learning algorithm that is automatically optimized according to different seasons and climate conditions.

[0011] Preferably, it further includes an automatic cleaning control step, when the cleanliness index is higher than a set threshold, the method automatically triggers the photovoltaic panel cleaning system to start, and controls the cleaning time and method according to the degree of contamination.

[0012] Preferably, the cleaning system includes at least one of a high-pressure water cleaning device, an ultrasonic cleaning device, a wind dust removal device, and an electrostatic dust removal device, and automatically selects a suitable cleaning method according to the type of pollution.

[0013] Preferably, the method further comprises the step of parallel detection of the surfaces of multiple photovoltaic panels, wherein the parallel detection is realized through a distributed computing architecture, the imaging data of the surfaces of multiple photovoltaic panels are analyzed simultaneously, and the detection efficiency is optimized through a task scheduling mechanism.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, pollution data on the surface of photovoltaic panels are collected through a variety of imaging devices, combined with adaptive filtering and morphological analysis, to achieve efficient extraction of pollution features, greatly improving the detection accuracy. The system can calculate the cleanliness index in real time, dynamically feedback the pollution situation, and output the detection results in time through the cloud platform or the Internet of Things interface to ensure the power generation efficiency of the photovoltaic panels. At the same time, the system can automatically trigger the cleaning device according to the cleanliness threshold, control the cleaning time and method, realize fully automated cleaning, and reduce the burden of manual maintenance. It is particularly suitable for the operation and maintenance management of large-scale photovoltaic fields. In addition, with the help of machine learning algorithms, the system can automatically adjust thresholds and parameters according to environmental monitoring data, historical records and seasonal changes to ensure effectiveness under different climatic conditions, and support parallel detection of multiple photovoltaic panels. Through distributed computing, the detection efficiency is optimized, the detection time is significantly reduced, and reliable data support is provided for subsequent operation and maintenance, ultimately improving the service life and power generation efficiency of photovoltaic panels. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention provides a flow chart of a photovoltaic panel cleanliness detection method. DETAILED DESCRIPTION

[0016] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0017] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0018] Implementation See also Figure 1 The present invention provides a technical solution: a method for detecting cleanliness of a photovoltaic panel, comprising the following steps: S1: Data acquisition, obtaining image data of the photovoltaic panel surface through imaging equipment installed around or above the photovoltaic panel, so as to comprehensively collect the pollution conditions on the photovoltaic panel surface; S2: Image preprocessing, preprocessing the image data, specifically including: Image denoising: remove random noise and light interference from images; Image grayscale transformation: convert image data from color space to grayscale image to facilitate subsequent feature extraction operations; Edge enhancement: Use the Sobel operator or Canny edge detection algorithm to perform edge enhancement on the grayscale image to highlight the contamination boundary features on the surface of the photovoltaic panel; S3: Pollution feature extraction, extracting pollution features on the surface of the photovoltaic panel from the preprocessed image, the pollution features include grayscale changes, light reflection unevenness, surface spot distribution, dust coverage area, etc., and feature extraction is achieved through morphological operation image analysis method; S4: cleanliness index calculation, based on the extracted pollution features, the cleanliness index of the photovoltaic panel is calculated, specifically, the various pollution features are weighted and summed through a preset weight model, and the result is normalized to the range of 0 to 1, where 0 represents complete cleanliness and 1 represents severe pollution; S5: Cleanliness judgment: compare the calculated cleanliness index with the preset cleanliness threshold. If the cleanliness index is higher than the threshold, it is judged that the photovoltaic panel needs to be cleaned and a cleaning prompt signal is output. If the cleanliness index is lower than the threshold, it is judged that no cleaning is required. S6: Data output: output the cleanliness index and cleaning judgment results to the user or automatic control system through the display screen, cloud platform or IoT interface to facilitate timely decision-making and cleaning operations; S7: Result recording and feedback: The test results are recorded in the database for subsequent analysis and cleaning plan formulation, and the parameter configuration of the imaging equipment is optimized through the feedback mechanism to improve the accuracy and efficiency of the test.

[0019] Furthermore, the imaging device includes at least one of a visible light camera, an infrared camera, an ultraviolet camera, and a multi-spectral sensor, which is used to obtain surface contamination conditions of photovoltaic panels within different spectral ranges.

[0020] In this embodiment, in actual applications, the pollution conditions on the surface of the photovoltaic panel may vary due to differences in regions, environments, seasons, etc. Therefore, the imaging device is not limited to a single type, but the most suitable device is selected according to actual needs.

[0021] Specifically, visible light cameras are used to capture visible pollution on the surface of photovoltaic panels, such as dust, bird droppings and other obvious pollutants. Infrared cameras can detect thermal differences that are invisible to the naked eye, especially at night or in cloudy conditions. Infrared cameras can provide additional pollution information to ensure all-weather monitoring. Ultraviolet cameras or multispectral sensors are suitable for capturing the characteristics of pollutants in different bands, especially in complex environments, such as deserts or near factories. Multispectral imaging can significantly improve the accuracy and breadth of pollution identification.

[0022] Furthermore, the image denoising step adopts an adaptive median filter or a Gaussian filter algorithm to dynamically adjust the filter parameters according to the noise characteristics of the environment where the photovoltaic panel is located to ensure image quality.

[0023] In this embodiment, image data is often affected by lighting, equipment noise, etc., and direct use may lead to reduced detection accuracy. Therefore, it is necessary to pre-process the collected images to improve image quality.

[0024] Specifically, the median filter or Gaussian filter algorithm can be dynamically adjusted according to the characteristics of the environmental noise. For example, in urban or industrial areas, the noise characteristics may be more complex, so more flexible filtering parameters are required. The color image is converted into a grayscale image to simplify data processing. Especially for scenes with high illumination differences, the grayscale transformation can better highlight the contrast characteristics of pollutants. The Sobel operator or Canny edge detection algorithm is used on the grayscale image to ensure that the boundaries of the pollutants are clearer, especially when the pollutants are small or unevenly distributed in the image, which helps to improve the accuracy of subsequent feature extraction.

[0025] Furthermore, the pollution feature extraction step is performed by a morphological analysis method, including applying morphological operations such as dilation, erosion, opening operation, closing operation, etc. to the grayscale image to extract the structural features and distribution features of the pollution area.

[0026] In this embodiment, according to the distribution of dust particles in different areas, the grayscale changes can be segmented and analyzed through a specific algorithm to obtain the area size of the polluted area. Since the coverage of pollutants will affect the reflective characteristics of the photovoltaic panel, the uneven changes in reflectivity can be used to detect which areas are severely polluted. Especially when the photovoltaic panel is tilted at a large angle or there are many obstructions, the reflectivity is an important reference indicator. Morphological operations (such as expansion, corrosion, opening and closing operations) are used to extract the shape, size, boundary and distribution of the polluted area, which is particularly effective for identifying particulate pollutants such as sand and dust.

[0027] Furthermore, the calculation of the cleanliness index is implemented through a weighted model, which includes a weighted combination of parameters such as grayscale change, surface brightness difference, spot distribution area, surface reflectivity, etc., and the weight parameters are adjusted according to the environmental characteristics of different regions.

[0028] In this embodiment, the cleanliness index is calculated using a preset weighted model based on the extracted pollution characteristics. Specifically, different pollution characteristics have different degrees of influence on the power generation efficiency of photovoltaic panels. Therefore, in the weighted model, the weight of each feature can be dynamically adjusted according to environmental factors (such as areas with frequent sandstorms, where the weight of dust characteristics is higher). The result of the weighted summation is normalized to the range of 0 to 1 to facilitate the intuitive representation of cleanliness judgment. The normalized result represents the overall cleanliness of the photovoltaic panel, with 0 indicating complete cleanliness and 1 indicating the most serious pollution.

[0029] Furthermore, the setting of the cleanliness threshold is dynamically adjusted according to historical cleaning data and environmental monitoring data, and the threshold setting method includes a machine learning algorithm, which is automatically optimized according to different seasons and climate conditions.

[0030] In this embodiment, the cleanliness judgment is achieved by comparing the calculated cleanliness index with a preset cleanliness threshold. The threshold can be dynamically adjusted based on historical cleaning records and environmental conditions. For example, the threshold can be optimized through a machine learning algorithm so that it can be automatically corrected according to different seasons and climatic conditions. During the long-term operation and maintenance of photovoltaic panels, the system can gradually accumulate data and establish a pollution prediction model so that cleaning plans can be made in advance.

[0031] Furthermore, when the cleanliness index is higher than a set threshold, the method automatically triggers the photovoltaic panel cleaning system to start, and controls the cleaning time and method according to the degree of contamination.

[0032] In this embodiment, the detection results are output to the operation and maintenance system through a display screen, a cloud platform or an Internet of Things interface, so that operators can view the cleaning status of the photovoltaic panels in real time. At the same time, the system has a linkage function with the automatic cleaning system. When the cleanliness index exceeds the threshold, the system can automatically trigger the cleaning equipment to start. The cleaning equipment can automatically adjust the cleaning strategy (such as water flow pressure, cleaning time, etc.) according to the type, area and severity of the pollution, thereby optimizing the cleaning effect.

[0033] Furthermore, the cleaning system includes at least one of a high-pressure water cleaning device, an ultrasonic cleaning device, a wind dust removal device, and an electrostatic dust removal device, and automatically selects a suitable cleaning method according to the type of pollution.

[0034] In this embodiment, the system supports dynamic optimization of the parameters and cleaning strategies of the imaging device through historical data. For example, the device can adjust the camera's shooting angle, lighting compensation and other settings based on the feedback results to improve detection accuracy and adapt to different environmental conditions. The system can also optimize the weighted model through the data feedback mechanism to further improve the accuracy of the cleanliness index calculation.

[0035] Furthermore, in the step of parallel detection of the surfaces of multiple photovoltaic panels, the parallel detection is achieved through a distributed computing architecture, the imaging data of the surfaces of multiple photovoltaic panels are analyzed simultaneously, and the detection efficiency is optimized through a task scheduling mechanism.

[0036] In this embodiment, in a large-scale photovoltaic power generation field, the imaging data of the surfaces of multiple photovoltaic panels can be processed simultaneously through a distributed computing architecture, and the task scheduling mechanism ensures the detection efficiency of multiple photovoltaic panels. Especially for large-area photovoltaic fields, it can significantly reduce the detection time while ensuring that the cleanliness status of each photovoltaic panel is accurately fed back. This architecture also supports distributed storage and processing of large-scale data, providing reliable data support for subsequent operation and maintenance plans.

[0037] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for detecting cleanliness of a photovoltaic panel, characterized in that: The following steps are involved: S1: Data acquisition, obtaining image data of the photovoltaic panel surface through imaging equipment installed around or above the photovoltaic panel, so as to comprehensively collect the pollution conditions on the photovoltaic panel surface; S2: Image preprocessing, preprocessing the image data, specifically including: Image denoising: remove random noise and light interference from images; Image grayscale transformation: convert image data from color space to grayscale image to facilitate subsequent feature extraction operations; Edge enhancement: Use the Sobel operator or Canny edge detection algorithm to perform edge enhancement on the grayscale image to highlight the contamination boundary features on the surface of the photovoltaic panel; S3: Pollution feature extraction, extracting pollution features on the surface of the photovoltaic panel from the preprocessed image, the pollution features include grayscale changes, light reflection unevenness, surface spot distribution, dust coverage area, etc., and feature extraction is achieved through morphological operation image analysis method; S4: cleanliness index calculation, based on the extracted pollution features, the cleanliness index of the photovoltaic panel is calculated, specifically, the various pollution features are weighted and summed through a preset weight model, and the result is normalized to the range of 0 to 1, where 0 represents complete cleanliness and 1 represents severe pollution; S5: Cleanliness judgment: compare the calculated cleanliness index with the preset cleanliness threshold. If the cleanliness index is higher than the threshold, it is judged that the photovoltaic panel needs to be cleaned and a cleaning prompt signal is output. If the cleanliness index is lower than the threshold, it is judged that no cleaning is required. S6: Data output: output the cleanliness index and cleaning judgment results to the user or the automated control system through a display screen, cloud platform or IoT interface to facilitate timely decision-making and cleaning operations; S7: Result recording and feedback: The test results are recorded in the database for subsequent analysis and cleaning plan formulation, and the parameter configuration of the imaging equipment is optimized through the feedback mechanism to improve the accuracy and efficiency of the test.

2. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The imaging device includes at least one of a visible light camera, an infrared camera, an ultraviolet camera, and a multi-spectral sensor, and is used to obtain surface contamination conditions of photovoltaic panels within different spectral ranges.

3. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The image denoising step adopts an adaptive median filter or a Gaussian filter algorithm, and dynamically adjusts the filter parameters according to the noise characteristics of the environment where the photovoltaic panel is located to ensure image quality.

4. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The pollution feature extraction step is performed by a morphological analysis method, including applying morphological operations such as expansion, corrosion, opening operation, closing operation, etc. to the grayscale image to extract the structural features and distribution features of the pollution area.

5. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The calculation of the cleanliness index is achieved through a weighted model, which includes a weighted combination of parameters such as grayscale change, surface brightness difference, spot distribution area, surface reflectivity, etc. The weight parameters are adjusted according to the environmental characteristics of different regions.

6. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The setting of the cleanliness threshold is dynamically adjusted according to historical cleaning data and environmental monitoring data. The setting method of the threshold includes a machine learning algorithm, which is automatically optimized according to different seasons and climate conditions.

7. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The method further includes an automatic cleaning control step. When the cleanliness index is higher than a set threshold, the method automatically triggers the photovoltaic panel cleaning system to start, and controls the cleaning time and method according to the degree of contamination.

8. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The cleaning system includes at least one of a high-pressure water cleaning device, an ultrasonic cleaning device, a wind dust removal device, and an electrostatic dust removal device, and automatically selects a suitable cleaning method according to the type of pollution.

9. The photovoltaic panel cleanliness detection method according to claim 1, characterized in that: The method further includes the step of parallel detection of multiple photovoltaic panel surfaces, wherein the parallel detection is achieved through a distributed computing architecture, imaging data of multiple photovoltaic panel surfaces are analyzed simultaneously, and detection efficiency is optimized through a task scheduling mechanism.

Images