Photovoltaic System Based on Transpiration Self-Cleaning

Abstract

The present invention discloses a photovoltaic system based on transpiration self-cleaning, which includes a whole machine frame and an automatic light-tracking mechanism, a height self-adaptive mechanism, a cleaning system, a sensing system and a control system arranged thereon. The automatic light-tracking mechanism can make the photovoltaic panel always face the sun signal. The height self-adaptive mechanism can adjust the vertical height of the whole machine frame, lower it close to the plants at night to improve the efficiency of transpiration, and raise it away from the plants during the day to avoid blocking sunlight. The cleaning system can transport the water collected by the plant transpiration to the photovoltaic panel on the base support plate and use the fine mist nozzles to achieve the purpose of self-cleaning of the photovoltaic panel. At the same time, the photovoltaic system of the present invention can also use the weighted average algorithm to integrate and analyze the sun position signals obtained by the vision system, the light intensity sensor and the temperature sensor for controlling the orientation of the photovoltaic panel to improve the efficiency of solar energy conversion. The present invention realizes the self-cleaning operation of the photovoltaic system based on the transpiration of plants, which not only reduces the cleaning and maintenance costs of the photovoltaic panel, but also improves the stability and long-term operation efficiency of the photovoltaic power generation system.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation, and specifically to a photovoltaic system based on transpiration self-cleaning. Background Art

[0002] As the core component of solar energy conversion, the photoelectric conversion efficiency of a photovoltaic panel directly determines the overall performance of a photovoltaic power generation system. However, a photovoltaic panel that is long-term exposed to the outdoor environment is vulnerable to various pollutants such as sand, bird droppings, and pollen, resulting in a layer of dirt covering its surface, thereby affecting the light transmittance and reducing the power generation efficiency.

[0003] Traditional cleaning methods for photovoltaic panels often rely on manual wiping or spraying with a high-pressure water pump. These methods are not only time-consuming and laborious, but also difficult to implement in water-scarce or arid regions, and even increase the waste of water resources. In some cases, although the use of chemical cleaners can remove stubborn stains, it may cause certain pollution to the environment, further increasing the cost and environmental burden of the cleaning work. Therefore, how to find an efficient, environmentally friendly and low-cost cleaning solution for photovoltaic panels has become an urgent problem in the industry.

[0004] In this context, plant transpiration, as a natural water cycle process widely existing in nature, demonstrates its potential in photovoltaic cleaning. Plants absorb water through their roots and discharge water through stomata to form water vapor. This process is not only part of the plant's life activities, but also provides moisture to the surrounding environment. Especially for plants planted near photovoltaic panels, the water vapor released by their transpiration can help clean the surface of the photovoltaic panels to a certain extent. This natural water cycle method is not only a low-cost and sustainable way to utilize water resources, but also can effectively reduce the dependence on traditional cleaning methods. Compared with traditional manual cleaning and chemical cleaning, the self-cleaning technology based on plant transpiration has multiple advantages such as reducing manual input, saving water resources, and reducing environmental pollution. It is an innovative solution that takes into account both economic benefits and environmental friendliness. The application of this technology can not only reduce the maintenance cost of photovoltaic panels, but also improve the stability and long-term operation efficiency of photovoltaic power generation systems. With the further research and promotion of this technology, future photovoltaic power stations may be able to form a more intelligent and environmentally friendly cleaning and maintenance mechanism through the synergistic effect with the surrounding vegetation, promoting the photovoltaic industry to develop towards a greener and more sustainable direction. Summary of the Invention

[0005] In order to solve the above problems, the present invention provides a photovoltaic system based on transpiration self-cleaning to achieve the purpose of efficient power generation of a solar photovoltaic panel.

[0006] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0007] A photovoltaic system based on transpiration self-cleaning, comprising an overall machine frame and an automatic light-tracking mechanism, a height self-adaptive mechanism, a cleaning system, a sensing system and a control system arranged thereon;

[0008] The overall machine frame includes an upper support shaft, a lower support shaft at the front side, a support shaft at the rear side and a base support plate. The automatic light-tracking mechanism is arranged on the base support plate, the height self-adaptive mechanism is arranged between the upper support shaft and the lower support shaft at the front side, and the control system is embedded inside the base support plate;

[0009] The automatic light-tracking mechanism includes a number of stepper motors to achieve the purpose of full-angle tracking of the sun signal by the photovoltaic panel thereon;

[0010] The height self-adaptive mechanism is used to adjust the vertical distance between the upper and lower support shafts at the front side of the overall machine frame according to the different heights of the plants below. The height self-adaptive mechanism includes a stop yoke assembly, a connecting rod assembly and a motor. An LED lamp is installed on the connecting rod assembly. The stop yoke assembly converts the circular motion provided by the motor into a motion in the height direction, and through the connecting rod assembly, it is converted into the adjustment of the vertical distance between the upper and lower support shafts;

[0011] The cleaning system includes a water pipe, a diaphragm pump, a water storage tank and a high-pressure fine mist nozzle. The base support plate stores the water collected by the transpiration of the plants into the water storage tank. The diaphragm pump transports the water in the water storage tank to the high-pressure fine mist nozzle through the water pipe to spray water for self-cleaning of the photovoltaic panel;

[0012] The sensing system includes a camera, a light intensity sensor and a temperature sensor, which are used to obtain the tracking data of the sun signal and transmit it to the control system in real time;

[0013] The control system includes an information acquisition system, a data processing system and a control system. The information acquisition system is used to obtain the sun signal tracking data transmitted by the sensing system in real time. The data processing system is used to process the sun signal tracking data to obtain the position coordinates of the sun relative to the photovoltaic panel. The control system controls the automatic light-tracking mechanism, the height self-adaptive mechanism and the cleaning system according to the position coordinates and control signals generated by the data processing system.

[0014] As a further improvement, the base support plate is in a shape that is high in the middle and low around, which is convenient for the dew condensation water to enter the water collection tank.

[0015] As a further improvement, the automatic light-tracking mechanism includes three stepper motors, which can realize the adjustment of six degrees of freedom of the photovoltaic panel thereon.

[0016] As a further improvement, the control system controls the rotation angle of the motor by sending different pulse numbers to the stepper motor.

[0017] As a further improvement, when the height adaptive mechanism rises, the LED lights on both sides are retracted to reduce the probability of damage to the mechanism; when the height adaptive mechanism descends, the LED lights on both sides are turned on, and the distance between the LED lights becomes larger, so that the illuminated area increases. Since the transpiration of plants is stimulated in an environment with high temperature and strong light, the height adaptive mechanism can improve the efficiency of plant transpiration.

[0018] As a further improvement, the whole machine frame is symmetrically distributed, and the center of gravity is located on the axis of the device to ensure the stability when the photovoltaic panel rotates.

[0019] As a further improvement, the upper and lower support shafts on the front side and the rear support shaft of the whole machine frame are made of aluminum profiles and are connected through connecting angle pieces and T-shaped nuts; the automatic light tracking mechanism, the height adaptive mechanism and the cleaning system are all connected and fastened to the whole machine frame by bolts.

[0020] As a further improvement, the anti-rotation yoke assembly of the height adaptive mechanism is connected to the upper part of the connecting rod assembly, and at the same time, the connecting rod assembly is designed to be lightweight according to the load condition.

[0021] As a further improvement, a ratchet structure is arranged in the anti-rotation yoke assembly of the height adaptive mechanism. The motor is not enabled and cannot be self-locked when it is not powered on. Therefore, when the motor is not powered on, the height adaptive mechanism will descend under its own weight, resulting in abnormal operation; the ratchet structure arranged in the anti-rotation yoke assembly enables the motor to be powered on and enabled only when it needs to work, and is not powered on usually, and relies on the self-locking of the ratchet to maintain the current position, so as to achieve the purpose of reducing power consumption.

[0022] As a further improvement, the cameras, light intensity sensors and temperature sensors of the sensing system are installed on both sides of the photovoltaic panel, and the angle of the camera is perpendicular to the plane of the photovoltaic panel.

[0023] As a further improvement, the control system includes a single-chip microcomputer, a communication module and a power module. The single-chip microcomputer uses a development board with the model of STM32F103, and the communication module is a USART module. The single-chip microcomputer can upload information in real time through the communication module and perform wireless remote control on each mechanism component of the device.

[0024] As a further improvement, the control system obtains the position coordinates of the sun relative to the photovoltaic panel based on the real-time tracking and detection method of the sun signal. The real-time tracking and detection method of the sun signal includes

[0025] Real-time capturing of the sun image by the camera, extraction of the position and morphological features of the sunspot through image processing, calculation of the azimuth angle and altitude angle of the sun signal relative to the photovoltaic panel using image processing algorithms, and obtaining the real-time azimuth information of the sun;

[0026] The change of solar radiation intensity is monitored in real time by a light intensity sensor. According to the light intensity data measured at different angles, the azimuth and altitude angle of the sun are deduced. At the same time, the change of the surface temperature of the photovoltaic panel is monitored in real time by a temperature sensor, which is used as an indirect index for auxiliary judgment of the sun's position.

[0027] Data fusion and tracking algorithm: By fusing the data of machine vision, light intensity sensor and temperature sensor, the weighted average algorithm is used to accurately calculate the position of the sun, and the calculated value of the final sun position coordinates is output.

[0028] As a further improvement, the data fusion and tracking algorithm includes

[0029] First, the azimuth, radiation intensity and temperature change of the sun are collected in real time through the machine vision system, light intensity sensor and temperature sensor, and three types of data are obtained: visual data (representing the estimated sun position output by the machine vision system), light intensity data (representing the estimated sun position output by the light intensity sensor), and temperature data (representing the estimated sun position output by the temperature sensor);

[0030] Then, the weight distribution is determined (weight of the machine vision system), (weight of the light intensity sensor), (weight of the temperature sensor). In the weighted average algorithm, the weight of each data source is used to measure the credibility or accuracy of the data. The selection of the weight can be based on the characteristics of the sensor, environmental conditions and the reliability of the real-time data;

[0031] Perform weighted average calculation, calculate the weighted data and output the final calculated value of the sun position.

[0032] As a further improvement, the formula of the weighted average algorithm is:

[0033]

[0034] Where is the calculated value of the finally estimated sun position, , and are the sun position data of the machine vision, light intensity sensor and temperature sensor respectively, , and are the weights of the above sensor data respectively.

[0035] As a further improvement, to cope with environmental changes and fluctuations in the data quality of different sensors, the weight allocation is determined through a dynamic weight adjustment mechanism:

[0036] When the machine vision system performs poorly under complex weather conditions (such as cloudy), its weight is dynamically reduced, and vice versa; the data quality of the light intensity sensor can be judged by the change range of the light intensity value. When the light intensity changes little or the sensor signal is unstable, its weight is reduced; the weight of the temperature sensor can be dynamically adjusted according to its correlation with the data of other sensors. When its data matches well with the data of other sensors, its weight can be increased.

[0037] As a further improvement, the image processing algorithm includes

[0038] Pre-collect images of the sun as the target object;

[0039] Use the Otsu method to binarize the collected image information, and then generate a binary image for convenient subsequent processing;

[0040] Annotate the boundary boxes of the sun in the collected images and assign a unique identifier to each mark. This process involves determining the length, width, and center coordinates of the boundary boxes to ensure that each sun target can be accurately identified and located in subsequent analysis;

[0041] Use Python and the OpenCV library to select regions of the image and equally divide the regions to achieve quantitative analysis of the sunlight intensity and direction; this step includes setting the height and width of the image matrix and determining the positions of the two vertices of the sun positioning region, and distinguishing different regions through color triples, thereby ensuring clarity visually and improving the recognition accuracy;

[0042] Save the processed image as a PNG format to ensure high quality during storage and transmission, facilitating subsequent analysis and model training;

[0043] Use the images with annotations and fixed region drawings to create a corresponding sun signal dataset and input it into a single-stage anchor-free deep convolutional neural network. In this step, through forward propagation and parameter update of the closed-loop system, an optimal pre-trained model is obtained to improve the accuracy of target detection and analysis;

[0044] During operation, the camera obtains video information of the sun target in the sky in real time, and this process reads the video stream frame by frame;

[0045] Input the data into the optimal pre-trained model to achieve quantitative analysis of the central position of the sun in the sky and the range of solar illumination intensity in the anchor box: Detect the solar target, locate its central position, and determine the range of illumination intensity in the anchor box, so as to obtain the position information of the sun relative to the photovoltaic panel.

[0046] As a further improvement, the quantitative analysis of the range of solar illumination intensity within the anchor box includes

[0047] Define as the quantization index of solar illumination intensity in every one-third of the fixed selection area, represent the overlapping area between the solar center recognition box and the fixed selection area in every one-third of the fixed area, represent the area of a single one-third fixed selection area; the specific formula for quantitative analysis of the solar illumination intensity within the anchor box is:

[0048]

[0049] Define three thresholds, threshold 1, threshold 2, and threshold 3. When exceeds these thresholds respectively, it is determined that the solar illumination intensity reaches mild, moderate, and severe levels;

[0050] The quantitative analysis formula for illumination intensity is as follows:

[0051]

[0052] Where represents the overlapping area between the target area and the light source irradiation area, represents the total area of the fixed selection area.

[0053] As a further improvement, in order to achieve the light chasing effect, it is necessary to adjust the position or direction of the target according to the position of the light source, which can be achieved by calculating the deviation between the center point of the target and the light source:

[0054]

[0055]

[0056] Where in the formula 、 represent the central position coordinates of the target, 、 represent the current coordinates of the light source;

[0057] During the light chasing process, continuously adjust the target position and track the light source position to form a closed-loop control to ensure the stability of the light chasing effect. The adjustment formula in this process is:

[0058]

[0059] in is the feedback coefficient to control the strength of the feedback.

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

[0061] 1. The present invention proposes a solar photovoltaic panel system with six degrees of freedom, which adopts three-axis technology to increase the light receiving time, improve the light receiving efficiency, and increase the light energy conversion efficiency. It also makes up for the deficiency of the existing technology that it cannot be extended in the vertical direction, and greatly improves the adaptability of the device.

[0062] 2. This invention achieves high-precision, real-time tracking of the sun's position by integrating data from machine vision, light intensity sensors, and temperature sensors using a dynamic weighted averaging algorithm. Compared to single-sensor solutions, this method offers greater environmental adaptability and maintains stable performance under complex conditions, including sunny, cloudy, and foggy conditions. A dynamic weight adjustment mechanism optimizes weight distribution based on the quality of real-time sensor data, effectively improving the system's robustness and accuracy, ensuring the device always faces the sun and increasing solar energy efficiency.

[0063] 3. The present invention has significant advantages such as saving water resources, reducing operating costs, and being environmentally friendly and sustainable. It achieves cleaning by utilizing the water produced by natural transpiration of plants, without the need for additional water sources or chemical detergents, and reducing dependence on labor, machinery and chemicals.

[0064] 4. This invention can be widely used in photovoltaic pastures to explore the organic combination of photovoltaic technology and animal husbandry. It can achieve multi-level animal husbandry, improve land utilization, and protect national security. Photovoltaic panels are installed at a specific height to form a sunshade, which can provide shade and rain protection for livestock in the pasture, and provide a good living environment for vegetation, protecting the ecological environment; it can also supplement light for pasture, improve pasture quality and yield, and superimpose the energy income brought by photovoltaic power generation, greatly increasing the economic output per unit of land;

[0065] Other beneficial effects of the present invention will be further described in the following specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0067] Figure 1 Schematic diagram of the structure of the photovoltaic system based on transpiration self-cleaning of the present invention;

[0068] Figure 2 This is a schematic diagram of the entire framework of the photovoltaic system based on transpiration self-cleaning of the present invention;

[0069] Figure 3Schematic diagram of the highly adaptive mechanism of the photovoltaic system based on transpiration self-cleaning of the present invention;

[0070] Figure 4 Schematic diagram of the automatic light-tracking mechanism of the photovoltaic system based on transpiration self-cleaning of the present invention;

[0071] Figure 5 Schematic diagram of the cleaning system of the photovoltaic system based on transpiration self-cleaning of the present invention (water pipes not shown);

[0072] Figure 6 Schematic diagram of the perception of the photovoltaic system based on transpiration self-cleaning of the present invention;

[0073] Figure 7 Working technical roadmap of the real-time solar signal tracking detection method of the present invention. Specific implementation manners

[0074] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0075] Embodiment 1

[0076] Referring to Figures 1 to 6 As shown, this embodiment provides a self-cleaning device for a solar photovoltaic panel, which includes an overall machine frame 1 and an automatic light-tracking mechanism 2, a height adaptive mechanism 3, a cleaning system 4, a perception system 5 and a control system arranged thereon. At the same time, it includes a photovoltaic panel 6 and an LED lamp 7.

[0077] The overall machine frame 1 integrates the automatic light-tracking mechanism 2, the height adaptive mechanism 3, the cleaning system 4, the perception system 5 and the control system. The overall machine frame 1 includes an upper support shaft 101 and a lower support shaft 102 at the front side, a support shaft 103 at the rear side, and a base support plate 104. The automatic light-tracking mechanism 2 is arranged on the base support plate 104. The height adaptive mechanism 3 is arranged between the upper support shaft 101 and the lower support shaft 102 at the front side. The control system is embedded inside the base support plate 104. The base support plate 104 is designed to be in a shape with a higher middle and lower peripheries, facilitating the dew condensation water to enter the water collection tank 403.

[0078] The automatic light-tracking mechanism 2 includes three stepper motors 201, 202 and 203. The control system controls the rotation angle of the motor by sending different pulse numbers to the stepper motor, enabling it to achieve six-degree-of-freedom adjustment, which is beneficial to adjusting the angle of the photovoltaic panel 6 arranged thereon to achieve the purpose that the photovoltaic panel 6 always faces the solar signal vertically.

[0079] The height adaptive mechanism 3 can adjust the vertical distance between the upper and lower support shafts 101 and 102 on the front side of the whole machine frame 1 according to the different heights of the plants planted under the LED lights 7, so as to change the height of the LED lights 7 to meet different height requirements and achieve the purpose of promoting plant transpiration. When the height adaptive mechanism 3 rises, the LED lights 7 placed on both sides move closer to the middle and fold up. Figure 3 Both ends of the shown link assembly are connected to the upper and lower support shafts 101 and 102. Therefore, when the height adaptive mechanism 3 rises, the two ends of the link assembly in the vertical direction are "stretched", and the distance between the two ends in the horizontal direction decreases, realizing the folding up of the LED lights 7 to reduce the probability of mechanism damage; when the height adaptive mechanism 3 descends, the LED lights 7 placed on both sides open, and the distance between the LED lights 7 becomes larger, so that the illuminated area increases. Since the transpiration of plants is stimulated in an environment with high temperature and strong light, the height adaptive mechanism 7 can improve the efficiency of plant transpiration. Specifically, the height adaptive mechanism 7 includes a rotation stopping yoke assembly 301, a link assembly 302 and a motor 303, and the LED lights 7 are arranged on the link assembly 302. The rotation stopping yoke assembly 301 converts the circular motion provided by the motor 303 into motion in the height direction, and then converts it into the adjustment of the vertical distance between the upper and lower support shafts 101 and 102 on the front side of the whole machine frame 1 through the link assembly 302: The rotation stopping yoke assembly 301 is fixed to the rear support shaft and its height remains unchanged all the time. When it makes a circular motion, it will drive the upper support shaft 101 to move up and down. Since the lower support shaft 102 is fixed to the ground, the distance between the upper and lower support shafts 101 and 102 changes accordingly, thus realizing the adjustment of the vertical distance, and further adjusting the height of the whole machine frame 1 to realize the height adjustment of the LED lights 7.

[0080] The cleaning system 4 is used to transport the moisture collected by the plant transpiration to the photovoltaic panel 6 on the base support plate 104 to achieve the purpose of self-cleaning of the photovoltaic panel 6. The cleaning system 4 includes a water pipe, a diaphragm pump 402, a water storage tank 403 and a high-pressure fine mist nozzle 401. The water vapor generated by the transpiration of plants condenses into dew on the base support plate 104, and the base support plate 104 stores the collected moisture in the water storage tank 403. The diaphragm pump 402 transports the moisture in the water storage tank 403 to the high-pressure fine mist nozzle 401 through the water pipe to spray and clean the photovoltaic panel 6, realizing the cleaning work of the photovoltaic panel 6.

[0081] The sensing system 5 includes a camera 501, a light intensity sensor 502 and a temperature sensor 503. The three obtain relevant data of the solar signal in real time and transmit it to the control system in real time.

[0082] The control system can control operations such as angle adjustment, height adjustment, and cleaning throughout the process of the photovoltaic system. The control system includes an information acquisition system, a data processing system, and a control system. The information acquisition system can obtain the sun signal tracking / sensing data transmitted by the sensing system in real time. The data processing system processes the relevant data of the sun signal to obtain the spatial position coordinates of the sun relative to the photovoltaic panel 6. The control system controls the automatic light tracking mechanism 2, the height adaptive mechanism 3, and the cleaning system 4 according to the position coordinates and control signals generated by the data processing system. At the same time, the control system is powered by the power generation of the photovoltaic panel 6.

[0083] In the embodiments of the present invention, the support shafts of the whole machine frame 1 are all made of aluminum profiles and are connected to the base through connecting angle pieces and T-shaped nuts. The automatic light tracking mechanism 2, the height adaptive mechanism 3, the cleaning system 4, and the sensing system 5 are all connected and fastened to the whole machine frame 1 by bolts. The control system is embedded inside the base support plate 104. The whole machine frame 1 is symmetrically distributed, and the center of gravity is located on the axis of the device to ensure the stability when the photovoltaic panel rotates.

[0084] In the embodiments of the present invention, the anti-rotation yoke assembly 301 of the height adaptive mechanism 3 is connected to the upper part of the link assembly 302, and at the same time, the link assembly 302 is designed for lightweight according to the load conditions.

[0085] A ratchet structure is provided in the anti-rotation yoke assembly 301 of the height adaptive mechanism 3. The motor 303 is not enabled and cannot be self-locked when it is not powered on. Therefore, when the motor 303 is not powered on, the height adaptive mechanism 3 will descend under its own weight, resulting in abnormal operation. The ratchet structure provided in the anti-rotation yoke assembly 301 enables the motor 303 to be powered on and enabled only when it needs to work, and is not powered on usually, and relies on the self-locking of the ratchet to maintain the current position, achieving the purpose of reducing power consumption.

[0086] In the embodiments of the present invention, the camera 501, the light intensity sensor 502, and the temperature sensor 503 of the sensing system 5 are installed on both sides of the photovoltaic panel 6, and the angle of the camera 501 is perpendicular to the plane of the photovoltaic panel 6. The camera 501 can monitor the sun signal within its field of view in real time and capture the corresponding real-time video information to transmit to the control system.

[0087] In the embodiments of the present invention, the control system includes a single-chip microcomputer, a communication module, and a power module. The single-chip microcomputer uses a development board with the model STM32F103, and the communication module is a USART module. The single-chip microcomputer can upload information in real time through the communication module and perform wireless remote control on each mechanism component of the device.

[0088] Embodiment 2

[0089] This embodiment relates to a real-time tracking and detection method for solar signals. The tracking and detection method can be applied to the control system of embodiment 1, so that it controls the photovoltaic system to achieve a tracking effect. Figure 7 The real-time tracking and detection method for solar signals includes:

[0090] The camera 501 captures the sun image in real time, extracts the position and morphological features of the sun spot through image processing, and uses the image processing algorithm to calculate the azimuth and altitude angle of the sun signal relative to the photovoltaic panel to obtain the real-time position information of the sun;

[0091] The light intensity sensor 502 monitors the changes in solar radiation intensity in real time, and the sun's azimuth and altitude are calculated based on the light intensity data measured at different angles. At the same time, the temperature sensor 503 monitors the changes in the surface temperature of the photovoltaic panel in real time as an indirect indicator to assist in determining the sun's position.

[0092] Data fusion and tracking algorithm: By fusing the data from machine vision, light intensity sensor 502 and temperature sensor 503, a weighted average algorithm is used to accurately calculate the position of the sun and output the final calculated value of the sun's position coordinates.

[0093] As an improvement to the above embodiment, the data fusion and tracking algorithm includes

[0094] First, the machine vision system, light intensity sensor 502, and temperature sensor 503 collect the sun's position, radiation intensity, and temperature changes in real time to obtain three types of data: visual data (represents the sun position estimate output by the machine vision system), light intensity data (representing the sun position estimate output by the light intensity sensor 502), temperature data (represents the sun position estimate output by the temperature sensor 503);

[0095] Then, determine the weight distribution (Weight of machine vision system), (weight of light intensity sensor 502), (Weight of temperature sensor 503). In the weighted average algorithm, the weight of each data source is used to measure the credibility or accuracy of the data. The selection of weights can be based on the characteristics of the sensor, environmental conditions, and the reliability of real-time data;

[0096] Perform weighted average calculation, calculate the weighted data and output the final calculated value of the sun's position.

[0097] The weighted average algorithm formula is:

[0098]

[0099] wherein is the calculated value of the final estimated solar position, , and are the solar position data of the machine vision, light intensity sensor, and temperature sensor respectively, , and are the weights of the above sensor data respectively.

[0100] As an improvement of the above embodiment, to cope with environmental changes and data quality fluctuations of different sensors, the weight allocation is determined through a dynamic weight adjustment mechanism:

[0101] When the machine vision system performs poorly under complex weather conditions (such as cloudy), its weight is dynamically reduced, otherwise its weight is increased; the data quality of the light intensity sensor 502 can be judged by the change range of the light intensity value. When the light intensity changes little or the sensor signal is unstable, its weight is reduced; the weight of the temperature sensor 503 can be dynamically adjusted according to its correlation with the data of other sensors. When its data matches well with the data of other sensors, its weight can be increased.

[0102] As an improvement of the above embodiment, the image processing algorithm includes

[0103] Pre-acquire images of the sun of the target object; pay attention to selecting images at different shooting angles and shooting time periods to collect image data, and screen the images to retain images with higher resolution;

[0104] Use the maximum inter-class variance method to binarize and segment the collected image information, and then generate a binary image for subsequent processing;

[0105] Annotate the boundary boxes of the sun in the collected images and assign a unique identifier to each marker. This process involves determining the length, width, and center coordinates of the boundary boxes to ensure that each solar target can be accurately identified and located in subsequent analysis;

[0106] Select regions of the image through Python and the OpenCV library, and equally divide the region to achieve quantitative analysis of the sunlight intensity and direction; this step includes setting the height and width of the image matrix, determining the positions of two vertices of the solar positioning region, and distinguishing different regions through color triples, so as to ensure clarity visually and improve the recognition accuracy;

[0107] Save the processed image as a PNG format to ensure that the image maintains high quality during storage and transmission, which is convenient for subsequent analysis and model training;

[0108] Generate a corresponding solar signal dataset using the images marked as completed and drawn with fixed regions, divide it proportionally into a training set, a test set, and a validation set, and input it into a single-stage anchor-free deep convolutional neural network. In this step, through forward propagation and parameter updates of the closed-loop system, obtain the optimal pre-trained YOLO model to improve the accuracy of object detection and analysis;

[0109] During operation, the camera continuously acquires video information of the solar target in the sky, and reads the video stream frame by frame;

[0110] Input the data into the optimal pre-trained model to achieve quantitative analysis of the central position of the sun in the sky and the range of solar illumination intensity in the anchor box: detect the solar target, locate its central position, and determine the range of illumination intensity in the anchor box, thereby obtaining the position information of the sun relative to the photovoltaic panel.

[0111] As an improvement of the above embodiment, the fixed region drawing method includes

[0112] Define the image as a matrix of size where is the height (number of rows) of the image, and is the width (number of columns) of the image; and respectively represent the two vertices of the fixed selection area for the solar positioning center; to distinguish the solar center recognition box and the fixed selection area, define the color as a triple indicating the values of blue ( ), green ( ), and red ( ), and the specific formula is as follows:

[0113]

[0114]

[0115]

[0116]

[0117]

[0118] The operation of drawing a rectangle of the fixed selection area for quantitative analysis of solar illumination intensity on the image matrix can be expressed as:

[0119]

[0120]

[0121] Subsequently, the fixed selected area of the drawn solar irradiance analysis is trisected and labeled as regions A, B, and C; among them, Indicates the boundary box thickness.

[0122] As an improvement to the above embodiment, the quantitative analysis of the solar irradiance range within the anchor box includes

[0123] Define As the solar irradiance quantization index in each one-third of the fixed selected area, Indicates the overlapping area between the solar center recognition box and the fixed selected area in each one-third of the fixed area, Indicates the area of a single one-third fixed selected area; the specific formula for quantitatively analyzing the solar irradiance within the anchor box is:

[0124]

[0125] Define three thresholds, threshold 1, threshold 2, and threshold 3. When Respectively exceed these thresholds, it is determined that the solar irradiance reaches mild, moderate, and severe;

[0126] The quantitative analysis formula for the irradiance is as follows:

[0127]

[0128] Where Indicates the overlapping area between the target area and the light source irradiation area, Indicates the total area of the fixed selected area.

[0129] As an improvement to the above embodiment, in order to achieve the light tracking effect, it is necessary to adjust the position or direction of the target according to the position of the solar light source, which can be achieved by calculating the deviation between the center point of the target and the solar light source:

[0130]

[0131]

[0132] Where in the formula 、 Indicate the center position coordinates of the target, 、 Indicate the current coordinates of the solar light source;

[0133] During the light tracking process, continuously adjust the target position and the tracking light source position to form a closed-loop control to ensure the stability of the light tracking effect. The adjustment formula in this process is:

[0134]

[0135] Among them is the feedback coefficient to control the intensity of feedback.

[0136] As an improvement of the above embodiment, when achieving the light chasing effect, it is necessary to continuously track the detected sun target and predict the position and speed of the target through a Kalman filter. The specific formula is:

[0137]

[0138]

[0139] Among them, in the formula is the target state, such as position, speed, etc.; is the state transition matrix, is the control input model, is the control input, is the process noise, is the observed value, is the observation matrix, is the observation noise.

[0140] Based on this, this embodiment can achieve the correction of the calculation deviation of the sun position coordinates. The sun position coordinates at the next moment obtained through Kalman filter prediction can be used as a reference or correction amount to adjust the actual measurement value and the tracking strategy of the system. For example:

[0141] Correct the current deviation: During the real-time tracking process, if there is a large deviation between the current measurement value (such as light intensity or visual data) and the Kalman filter prediction value, this deviation can be applied to the correction of the current coordinates to improve the tracking accuracy.

[0142] Predictive error correction: Use the prediction result provided by the Kalman filter as a reference benchmark. If the real-time measured sun position deviates from the predicted value by more than a certain threshold, the system is triggered to make adjustments or recalibrations.

[0143] Utilize the prediction ability of the Kalman filter to "guide" the current tracking path, enabling the solar photovoltaic system to respond to changes in the sun position in advance and reducing the error accumulation caused by sensor noise or environmental changes.

[0144] Embodiment 3

[0145] Under the action of the sun signal real-time tracking detection method, this embodiment provides a working method for a self-cleaning device of a solar photovoltaic panel, including:

[0146] Step S1: Start the height adaptive device 3 and adjust the height of the led lamp 7 according to the height of the plants below the led lamp 7 to better promote the transpiration of the plants;

[0147] Step S2: Activate the sensing system 5, and obtain the solar signal information in real time through the real-time tracking and detection method of the solar signal. When the position information of the solar signal deviates from the center position of the photovoltaic panel 6, transmit a signal to the control system through the serial port protocol;

[0148] Step S3: The automatic light-tracking mechanism 2 adjusts the angle of the photovoltaic panel 6 according to the control signal, so that the photovoltaic panel 6 is always perpendicular to the solar signal, in order to achieve better power generation efficiency;

[0149] Step S4: While the solar photovoltaic panel 6 is working, the cleaning system 4 transports the water collected by the transpiration of plants to the high-pressure fine mist nozzle 401 through the diaphragm pump 402 to complete the self-cleaning work of the solar photovoltaic panel 6.

[0150] It should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A photovoltaic system based on transpiration self-cleaning, characterized in that: including a whole machine frame and an automatic light tracking mechanism, a height self - adaptation mechanism, a cleaning system, a sensing system and a control system arranged thereon; the whole machine frame includes several support shafts and a base support plate, the automatic light tracking mechanism is arranged on the base support plate, the height self - adaptation mechanism is arranged between the support shafts at the front side, and the control system is embedded inside the base support plate; the automatic light tracking mechanism includes several stepper motors to achieve the purpose of full - angle tracking of the sun signal by the photovoltaic panel thereon; the height self - adaptation mechanism is used to adjust the vertical distance of the support shafts on the front side of the whole machine frame. The height self - adaptation mechanism includes a non - rotating yoke assembly, a connecting rod assembly and a motor. A light source is installed on the connecting rod assembly. The non - rotating yoke assembly converts the circular motion provided by the motor into a motion in the height direction, and through the connecting rod assembly, it is converted into the adjustment of the vertical distance of the support shafts; the cleaning system includes a water pipe, a diaphragm pump, a water storage tank and a high - pressure fine mist nozzle. The base support plate stores the water collected due to plant transpiration into the water storage tank. The diaphragm pump transports the water in the water storage tank through the water pipe to the high - pressure fine mist nozzle to spray and self - clean the photovoltaic panel; the sensing system includes a camera, a light intensity sensor and a temperature sensor, which are used to obtain the tracking data of the sun signal and transmit it to the control system in real time; the control system includes an information acquisition system, a data processing system and a control system. The information acquisition system is used to obtain the sun signal tracking data transmitted by the sensing system in real time. The data processing system is used to process the sun signal tracking data to obtain the position coordinates of the sun relative to the photovoltaic panel. The control system controls the automatic light tracking mechanism, the height self - adaptation mechanism and the cleaning system according to the position coordinates and control signals generated by the data processing system.

2. The transpiration - based self - cleaning photovoltaic system according to claim 1, wherein: when the height self - adaptation mechanism rises, the light sources placed on both sides are retracted to reduce the probability of mechanism damage; when the height self - adaptation mechanism drops, the light sources placed on both sides are turned on, and the distance between the light sources on both sides becomes larger, so that the illuminated area increases. Since plant transpiration is stimulated in an environment with high temperature and strong light, the height self - adaptation mechanism can improve the efficiency of plant transpiration.

3. The transpiration - based self - cleaning photovoltaic system according to claim 1, wherein: a ratchet structure is arranged in the non - rotating yoke assembly of the height self - adaptation mechanism, so that the motor is only powered on and enabled when it needs to work, and is not powered on usually, and relies on the self - locking of the ratchet to maintain the current position, achieving the purpose of reducing power consumption.

4. The transpiration - based self - cleaning photovoltaic system according to claim 1, wherein: the control system obtains the position coordinates of the sun relative to the photovoltaic panel based on the real - time tracking and detection method of the sun signal. The real - time tracking and detection method of the sun signal includes capturing the sun image in real time through the camera, extracting the position and morphological features of the sun spot through image processing, calculating the azimuth angle and elevation angle of the sun signal relative to the photovoltaic panel by using the image processing algorithm, and obtaining the real - time azimuth information of the sun; The change of solar radiation intensity is monitored in real time by a light intensity sensor. According to the light intensity data measured at different angles, the azimuth and altitude angle of the sun are deduced. At the same time, the change of the surface temperature of the photovoltaic panel is monitored in real time by a temperature sensor, which is used as an indirect index for auxiliary judgment of the sun's position. Data fusion and tracking algorithm: By fusing the data of machine vision, light intensity sensor and temperature sensor, the weighted average algorithm is used to accurately calculate the position of the sun, and the calculated value of the final sun position coordinates is output.

5. The photovoltaic system based on transpiration self-cleaning according to claim 4, characterized in that: The data fusion and tracking algorithm includes First, the machine vision system, light intensity sensor, and temperature sensor are used to collect the azimuth, radiation intensity, and temperature changes of the sun in real time, obtaining three types of data: visual data , light intensity data , and temperature data ; Then, determine the weight distribution: the weight of the machine vision system , the weight of the light intensity sensor , the weight of the temperature sensor , perform a weighted average calculation, calculate the weighted data, and output the final calculated value of the sun position.

6. The photovoltaic system based on transpiration self-cleaning according to claim 5, characterized in that: To cope with environmental changes and data quality fluctuations of different sensors, the weight distribution is determined through a dynamic weight adjustment mechanism: When the machine vision system performs poorly under complex weather conditions, its weight is dynamically reduced, and vice versa; the data quality of the light intensity sensor can be judged by the change range of the light intensity value. When the light intensity changes little or the sensor signal is unstable, its weight is reduced; the weight of the temperature sensor can be dynamically adjusted according to its correlation with the data of other sensors. When its data matches well with the data of other sensors, its weight can be increased.

7. The photovoltaic system based on transpiration self-cleaning according to claim 4, characterized in that: The image processing algorithm includes Pre-collect images of the target object, the sun; The collected image information is binarized by the maximum inter-class variance method to generate a binary image for subsequent processing; The sun in the collected image is marked with a bounding box, and a unique identifier is assigned to each mark. This process involves determining the length, width and center coordinates of the bounding box to ensure accurate identification and positioning of each sun target in subsequent analysis; The image is selected by region through Python and the OpenCV library, and the region is equally divided to realize the quantitative analysis of the sunlight intensity and direction; this step includes setting the height and width of the image matrix and determining the positions of the two vertices of the sun positioning region, and distinguishing different regions through color triples, so as to ensure clarity visually and improve the recognition accuracy; The processed image is saved as a PNG format to ensure high quality of the image during storage and transmission, which is convenient for subsequent analysis and model training; The corresponding sun signal data set is made using the image with annotation and fixed region drawing completed, and it is input into a single-stage anchor-free deep convolutional neural network. In this step, through forward propagation and parameter update of the closed-loop system, an optimal pre-trained model is obtained to improve the accuracy of target detection and analysis; During operation, the camera obtains video information of the sun target in the sky in real time, and this process reads the video stream frame by frame; The data is input into the optimal pre-trained model to realize the quantitative analysis of the central position of the sun in the sky and the range of sunlight intensity in the anchor box: the sun target is detected, its central position is located, and the range of light intensity in the anchor box is judged, so as to obtain the position information of the sun relative to the photovoltaic panel.

8. The photovoltaic system based on transpiration self-cleaning according to claim 7, characterized in that: The quantitative analysis of the solar light intensity range within the anchor frame includes Definition is the quantization index of the solar light intensity in each one-third fixed selection area, represents the overlapping area between the solar center recognition frame and the fixed selection area in each one-third fixed area, represents the area of a single one-third fixed selection area; the specific formula for quantitatively analyzing the solar light intensity within the anchor box is: Define three thresholds, threshold 1, threshold 2, and threshold 3. When exceed these thresholds respectively, it is determined that the solar light intensity reaches mild, moderate, and severe levels.

9. The photovoltaic system based on transpiration self-cleaning according to claim 4, characterized in that: In order to achieve the light chasing effect, it is necessary to adjust the position or direction of the target according to the position of the light source, which can be achieved by calculating the deviation between the center point of the target and the light source: Among them, in the formula , represent the central position coordinates of the target, , represent the current coordinates of the light source.

10. The photovoltaic system based on transpiration self-cleaning according to claim 9, characterized in that: During the light chasing process, continuously adjust the target position and track the light source position to form a closed-loop control to ensure the stability of the light chasing effect. The adjustment formula in this process is: wherein is the feedback coefficient to control the intensity of the feedback.

Images