Photovoltaic system based on transpiration self-cleaning

Abstract

The invention discloses a photovoltaic system based on transpiration self-cleaning, which comprises a complete machine frame, and an automatic light following mechanism, a height self-adaptive mechanism, a cleaning system, a sensing system and a control system which are arranged on the complete machine frame, the height self-adaption mechanism can adjust the vertical height of the whole machine frame, the height self-adaption mechanism is lowered to be close to plants at night so as to improve the transpiration efficiency, and the height self-adaption mechanism is raised to be away from the plants in the daytime so as to avoid sunshine shielding, and the cleaning system can transport water collected by the plant transpiration on the base supporting plate to the photovoltaic panel. And the purpose of self-cleaning of the photovoltaic panel is achieved by using the mist spray head. Meanwhile, according to the photovoltaic system disclosed by the invention, sun position signals acquired by the visual system, the light intensity sensor and the temperature sensor can be integrated and analyzed by utilizing a weighted average algorithm, so that the orientation of the photovoltaic panel is controlled, and the solar energy conversion efficiency is improved. The self-cleaning operation of the photovoltaic system is realized based on the transpiration of plants, the cleaning and maintenance cost of the photovoltaic panel is reduced, and the stability and long-term operation efficiency of the photovoltaic power generation system are improved.

Description

Technical Field

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

[0002] Photovoltaic panels are the core components of solar energy conversion, and their photoelectric conversion efficiency directly determines the overall performance of the photovoltaic power generation system. However, photovoltaic panels exposed to the outdoor environment for a long time are easily invaded by various pollutants such as wind and sand, bird droppings, and pollen, resulting in a layer of dirt on the surface, which in turn affects the light transmittance and reduces the power generation efficiency.

[0003] Traditional photovoltaic panel cleaning methods often rely on manual wiping or the use of high-pressure water pumps to spray and clean. These methods are not only time-consuming and labor-intensive, but also difficult to implement in water-scarce or drought-stricken areas, and may even increase water waste. In some cases, although the use of chemical cleaners can remove stubborn stains, they may cause certain pollution to the environment, further increasing the cost and environmental burden of cleaning work. Therefore, how to find an efficient, environmentally friendly and low-cost photovoltaic panel cleaning solution has become an urgent problem to be solved in the industry.

[0004] In this context, plant transpiration, as a water cycle process that is ubiquitous in nature, shows its potential in photovoltaic cleaning. Plants absorb water through their roots and discharge water through their stomata to form water vapor. This process is not only part of plant life activities, but also provides water to the surrounding environment. In particular, the water vapor released by the transpiration of plants planted near photovoltaic panels can help clean the surface of photovoltaic panels to a certain extent. This natural water cycle is not only a low-cost and sustainable way to utilize water resources, but also effectively reduces dependence on traditional cleaning methods. Compared with traditional manual cleaning and chemical cleaning, self-cleaning technology based on plant transpiration has multiple advantages such as reducing labor 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 synergy with surrounding vegetation, and promote the photovoltaic industry to develop in a more green and 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 solar photovoltaic panels.

[0006] The technical solution adopted by the present invention to solve the technical problem is: A photovoltaic system based on transpiration self-cleaning, including a whole machine frame and an automatic light-chasing mechanism, a highly adaptive mechanism, a cleaning system, a sensing system and a control system arranged thereon; The whole machine frame includes an upper support shaft, a lower support shaft, a rear support shaft and a base support plate on the front side, the automatic light tracking mechanism is arranged on the base support plate, the height adaptive mechanism is arranged between the upper support shaft and the lower support shaft on the front side, and the control system is embedded in the base support plate; The automatic light-tracking mechanism includes a plurality of stepper motors to achieve the purpose of full-angle tracking of solar signals by the photovoltaic panels thereon; The height adaptive mechanism is used to adjust the vertical distance between the upper and lower support shafts on the front side of the whole machine frame according to the different heights of the plants below. The height adaptive mechanism includes a scotch yoke assembly, a connecting rod assembly and a motor. The connecting rod assembly is equipped with an LED lamp. The scotch yoke assembly converts the circular motion provided by the motor into a motion in the height direction, which is converted into the adjustment of the vertical distance between the upper and lower support shafts through the connecting rod assembly. The cleaning system includes a water pipe, a diaphragm pump, a water tank and a high-pressure fine mist nozzle. The base support plate stores the water collected by plant transpiration in the water tank. The diaphragm pump transports the water in the water tank to the high-pressure fine mist nozzle through the water pipe to spray water on the photovoltaic panel for self-cleaning. The sensing system includes a camera, a light intensity sensor and a temperature sensor, which are used to obtain tracking data of the solar signal and transmit it to the control system in real time; The control system includes an information collection system, a data processing system and a control system. The information collection system is used to obtain the solar signal tracking data transmitted by the perception system in real time. The data processing system is used to process the solar 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 highly adaptive mechanism and the cleaning system according to the position coordinates and control signals generated by the data processing system.

[0007] As a further improvement, the base support plate is high in the middle and low around, so as to facilitate the condensed water to enter the water collection tank.

[0008] As a further improvement, the automatic light-chasing mechanism includes three stepper motors, which can achieve six degrees of freedom adjustment of the photovoltaic panel thereon.

[0009] As a further improvement, the control system achieves the purpose of controlling the rotation angle of the motor by sending different numbers of pulses to the stepper motor.

[0010] As a further improvement, when the height adaptive mechanism is raised, the LED lights on both sides are retracted to reduce the chance of damage to the mechanism; when the height adaptive mechanism is lowered, the LED lights on both sides are turned on, and the distance between the LED lights becomes larger, thereby increasing the illuminated area. Since plant transpiration is stimulated in an environment of high temperature and strong light, the height adaptive mechanism can improve the efficiency of plant transpiration.

[0011] As a further improvement, the entire machine frame is symmetrically distributed with the center of gravity located on the axis of the device to ensure the stability of the photovoltaic panel during rotation.

[0012] 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 by connecting angle pieces and T-nuts; the automatic tracking mechanism, height adaptive mechanism and cleaning system are all connected and fastened to the whole machine frame by bolts.

[0013] As a further improvement, the yoke assembly of the highly adaptive mechanism is connected to the upper part of the connecting rod assembly, and the connecting rod assembly is designed to be lightweight according to the load conditions.

[0014] As a further improvement, a ratchet structure is provided in the anti-swivel yoke assembly of the highly adaptive mechanism, so that the motor is not enabled and cannot self-lock when no power is supplied. Therefore, when the motor is not powered, the highly adaptive mechanism will drop under the action of its own weight, resulting in malfunctioning of the mechanism. The ratchet structure provided in the anti-swivel yoke assembly enables the motor to be powered on and enabled only when it needs to work. It is not powered on at other times and relies on the self-locking of the ratchet to maintain the current position, thereby achieving the purpose of reducing power consumption.

[0015] As a further improvement, the camera, light intensity sensor and temperature sensor of the perception system are installed on both sides of the photovoltaic panel, wherein the camera angle is perpendicular to the plane of the photovoltaic panel.

[0016] As a further improvement, the control system includes a single-chip microcomputer, a communication module and a power module, the single-chip microcomputer adopts a development board of 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 various mechanism components of the device.

[0017] 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 detection method of the solar signal, and the real-time tracking detection method of the solar signal includes: The camera captures the sun image in real time, extracts the position and morphological characteristics 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; The light intensity sensor monitors the changes in solar radiation intensity in real time, and calculates the sun's azimuth and altitude angle based on the light intensity data measured at different angles. At the same time, the temperature sensor 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. Data fusion and tracking algorithm: By fusing the data from machine vision, light intensity sensor and temperature sensor, 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.

[0018] As a further improvement, the data fusion and tracking algorithm includes First, the machine vision system, light intensity sensor, and temperature sensor 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 (represents the sun's position estimate output by the light intensity sensor), temperature data (represents the sun position estimate output by the temperature sensor); Then, determine the weight distribution (Weight of machine vision system), (weight of light intensity sensor), (Weight of 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 weight can be based on the characteristics of the sensor, environmental conditions, and the reliability of real-time data; Perform weighted average calculation, calculate weighted data and output the final calculated value of the sun's position.

[0019] As a further improvement, the weighted average algorithm formula is:

[0020] in is the final estimated calculated value of the sun's position, , and They are the sun position data from machine vision, light intensity sensor, and temperature sensor. , and are the weights of the above sensor data respectively.

[0021] As a further improvement, in order to cope with environmental changes and fluctuations in data quality from different sensors, a dynamic weight adjustment mechanism is used to determine the weight distribution: When the machine vision system performs poorly under complex weather conditions (such as cloudy weather), its weight is dynamically reduced, and vice versa. The data quality of the light intensity sensor can be judged by the change in the light intensity value. When the light intensity change is small 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 other sensor data. When its data matches well with other sensor data, its weight can be increased.

[0022] As a further improvement, the image processing algorithm includes Pre-collecting an image of the target object, the sun; The collected image information is subjected to binary segmentation using the maximum inter-class variance method, thereby generating a binary image for subsequent processing; Annotate the sun in the acquired images with bounding boxes and assign a unique identifier to each marker. This process involves determining the length, width, and center coordinates of the bounding box to ensure that each sun target can be accurately identified and located in subsequent analysis. The image is selected and divided into equal parts using Python and OpenCV libraries to achieve quantitative analysis of the intensity and direction of sunlight. This step includes setting the height and width of the image matrix and determining the two vertex positions of the sun positioning area. Different areas are distinguished by color triplets, thereby visually ensuring clarity and improving recognition accuracy. Convert the processed images to PNG format to ensure that the images maintain high quality during storage and transmission, which is convenient for subsequent analysis and model training. The corresponding solar signal dataset is created using the images with completed annotations and fixed area drawing, and is passed into a single-stage anchor-free deep convolutional neural network. In this step, the optimal pre-training model is obtained through forward propagation and parameter update of the closed-loop system to improve the accuracy of target detection and analysis. When working, the camera obtains video information of the sun target in the sky in real time, and the process reads the video stream frame by frame; The data is input into the optimal pre-trained model to achieve quantitative analysis of the center position of the sun in the sky and the range of solar light intensity in the anchor frame: the solar target is detected, its center position is located, and the range of light intensity in the anchor frame is determined to obtain the position information of the sun relative to the photovoltaic panel.

[0023] As a further improvement, the quantitative analysis of the range of sunlight intensity within the anchor frame includes definition A quantitative index of the solar intensity in each third of the fixed selected area, It represents the overlapping area of ​​the sun center identification box and the fixed selection area in every one-third of the fixed area. represents the area of ​​a single one-third fixed selection area; the specific formula for quantitative analysis of the sunlight intensity in the anchor frame is:

[0024] Define three thresholds, Threshold 1, Threshold 2, and Threshold 3. When these thresholds are exceeded, the solar radiation intensity is judged to be mild, moderate, and severe; The quantitative analysis formula of light intensity is as follows:

[0025] in Represents the overlapping area of ​​the target area and the light source illumination area, Represents the total area of ​​the fixed selection region.

[0026] 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. This can be achieved by calculating the deviation between the center point of the target and the light source:

[0027]

[0028] In the formula , Indicates the center position coordinates of the target, , Indicates the current coordinates of the light source; In the process of light chasing, the target position and the position of the tracking light source are continuously adjusted to form a closed-loop control to ensure the stability of the light chasing effect. The adjustment formula in this process is:

[0029] in is the feedback coefficient to control the strength of the feedback.

[0030] Compared with the prior art, the present invention has the following beneficial effects: 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, increase the light energy conversion efficiency, and make up for the deficiency of the prior art that it cannot be extended in the vertical direction, thereby greatly improving the adaptability of the device.

[0031] 2. The present invention integrates the data of machine vision, light intensity sensor and temperature sensor, and uses a dynamic weighted average algorithm to achieve high-precision real-time tracking of the sun's position. Compared with a single sensor solution, this method has stronger environmental adaptability and can maintain stable performance under complex conditions such as sunny, cloudy, and haze. The dynamic weight adjustment mechanism optimizes the weight distribution according to the real-time data quality of the sensor, effectively improving the robustness and accuracy of the system, ensuring that the device is always facing the sun, and improving the efficiency of solar energy utilization.

[0032] 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 cleaning agents, and reducing dependence on labor, machinery, and chemical products.

[0033] 4. The present 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 safeguard national security. Photovoltaic panels are installed at a specific height to form a sunshade, which can provide a sunshade and rainproof environment for livestock in the pasture, and provide a good living environment for vegetation, protecting the ecological environment; it can supplement light for pasture, improve pasture quality and yield, and superimpose with the energy income brought by photovoltaic power generation, greatly increasing the economic output per unit of land; Other beneficial effects of the present invention will be further described in the following specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments: Figure 1 It is a schematic structural diagram of a photovoltaic system based on transpiration self-cleaning of the present invention; Figure 2 It is a schematic diagram of the whole machine framework of the photovoltaic system based on transpiration self-cleaning of the present invention; Figure 3 A schematic diagram of a highly adaptive mechanism of a photovoltaic system based on transpiration self-cleaning of the present invention; Figure 4 A schematic diagram of an automatic light-chasing mechanism of a photovoltaic system based on transpiration self-cleaning of the present invention; Figure 5 A schematic diagram of a cleaning system of a photovoltaic system based on transpiration self-cleaning according to the present invention (water pipes are not shown); Figure 6 A sensing schematic diagram of a photovoltaic system based on transpiration self-cleaning of the present invention; Figure 7 It is the working technical roadmap of the real-time tracking and detection method of solar signal of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Example 1 Reference Figures 1 to 6 As shown, this embodiment provides a solar photovoltaic panel self-cleaning device, which includes a whole machine frame 1 and an automatic light-chasing mechanism 2, a highly adaptive mechanism 3, a cleaning system 4, a sensing system 5 and a control system arranged thereon, and also includes a photovoltaic panel 6 and an LED lamp 7.

[0037] The whole machine frame 1 integrates the automatic light tracking mechanism 2, the height adaptive mechanism 3, the cleaning system 4, the sensing system 5 and the control system. The whole machine frame 1 includes an upper support shaft 101, a lower support shaft 102, a rear support shaft 103 and a base support plate 104 on the front side. 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 on the front side, and the control system is embedded in the base support plate 104. The base support plate 104 is designed to be high in the middle and low around, so as to facilitate the condensed water to enter the water collection tank 403.

[0038] The automatic light-chasing mechanism 2 includes three stepper motors 201, 202 and 203. The control system controls the rotation angle of the motor by sending different numbers of pulses to the stepper motor, so that it can be adjusted in six degrees of freedom, which is conducive to adjusting the angle of the photovoltaic panel 6 arranged thereon, so as to achieve the purpose of the photovoltaic panel 6 always facing the solar signal vertically.

[0039] 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 frame 1 according to the different heights of the plants planted under the LED lamps 7, thereby changing the height of the LED lamps 7 to adapt to different height requirements and achieve the purpose of promoting plant transpiration. When the height adaptive mechanism 3 is raised, the LED lamps 7 placed on both sides are retracted towards the middle. Figure 3The two ends of the connecting rod assembly are connected to the upper and lower support shafts 101 and 102, so when the height adaptive mechanism 3 is raised, the two ends of the connecting rod assembly in the vertical direction are "stretched", and the distance between the two ends in the horizontal direction is reduced, so that the LED lamp 7 is retracted to reduce the probability of damage to the mechanism; when the height adaptive mechanism 3 is lowered, the LED lamps 7 placed on both sides are turned on, and the distance between the LED lamps 7 becomes larger, so that the illuminated area is increased. 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 yoke assembly 301, a connecting rod assembly 302 and a motor 303, and the LED lamp 7 is arranged on the connecting rod assembly 302. The stop 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 entire machine frame 1 through the connecting rod assembly 302: the stop yoke assembly 301 is fixed to the rear support shaft, and the height remains unchanged. When it performs 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 is changed, thereby realizing the adjustment of the vertical distance, and then adjusting the height of the entire machine frame 1, thereby realizing the height adjustment of the LED lamp 7.

[0040] The cleaning system 4 is used to transport the water collected by the plant transpiration on the base support plate 104 to the photovoltaic panel 6, so as 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 tank 403 and a high-pressure fine mist nozzle 401. The water vapor generated by the transpiration of the plant condenses on the base support plate 104 when it is cooled. The base support plate 104 stores the collected water in the water tank 403. The diaphragm pump 402 transports the water in the water tank 403 to the high-pressure fine mist nozzle 401 through the water pipe, sprays water on the photovoltaic panel 6, and cleans the photovoltaic panel 6.

[0041] The sensing system 5 includes a camera 501, a light intensity sensor 502 and a temperature sensor 503, which acquire relevant data of the solar signal in real time and transmit it to the control system in real time.

[0042] The control system can control the angle adjustment, height adjustment and cleaning operations of the photovoltaic system throughout the whole process. The control system includes an information acquisition system, a data processing system and a control system. The information acquisition system can obtain the solar signal tracking / sensing data transmitted by the perception system in real time, and the data processing system processes the relevant data of the solar 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 photovoltaic panel 6.

[0043] In the embodiment of the present invention, the support shafts of the whole machine frame 1 are made of aluminum profiles, 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 through bolts, and the control system is embedded in the base support plate 104. The whole machine frame 1 adopts a symmetrical distribution, and the center of gravity is located at the device axis to ensure the stability of the photovoltaic panel when rotating.

[0044] In the embodiment of the present invention, the yoke assembly 301 of the height adaptive mechanism 3 is connected to the upper part of the connecting rod assembly 302, and the connecting rod assembly 302 is designed to be lightweight according to the load conditions.

[0045] A ratchet structure is provided in the stop yoke assembly 301 of the highly adaptive mechanism 3, and the motor 303 is not enabled when no power is supplied, and cannot self-lock. Therefore, when the motor 303 is not powered, the highly adaptive mechanism 3 will drop under the action of its own weight, resulting in malfunction of the mechanism. The ratchet structure provided in the stop yoke assembly 301 enables the motor 303 to be powered on and enabled only when it needs to work, and is not powered on at other times, relying on the self-locking of the ratchet to maintain the current position, thereby achieving the purpose of reducing power consumption.

[0046] In the embodiment of the present invention, the camera 501, light intensity sensor 502 and temperature sensor 503 of the sensing system 5 are installed on both sides of the photovoltaic panel 6, wherein the angle of the camera 501 is perpendicular to the plane of the photovoltaic panel 6. The camera 501 can monitor the solar signal within its field of view in real time and capture the corresponding real-time video information for transmission to the control system.

[0047] In the embodiment of the present invention, the control system includes a single-chip microcomputer, a communication module and a power module, the single-chip microcomputer adopts a development board of 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 various mechanism components of the device.

[0048] Example 2 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 light-tracking effect. Figure 7 The real-time tracking and detection method of solar signals includes: The camera 501 is used to capture the sun image in real time, and the position and morphological characteristics of the sun spot are extracted through image processing. The azimuth and altitude angle of the sun signal relative to the photovoltaic panel are calculated using the image processing algorithm to obtain the real-time azimuth information of the sun; The light intensity sensor 502 monitors the change of solar radiation intensity in real time, and calculates the azimuth and altitude of the sun based on the light intensity data measured at different angles; at the same time, the temperature sensor 503 monitors the change of the surface temperature of the photovoltaic panel in real time as an indirect indicator to assist in determining the position of the sun; Data fusion and tracking algorithm: By fusing the data of 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.

[0049] As an improvement of the above embodiment, the data fusion and tracking algorithm includes 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); 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, and the selection of weight can be based on the characteristics of the sensor, environmental conditions, and the reliability of real-time data; Perform weighted average calculation, calculate weighted data and output the final calculated value of the sun's position.

[0050] The weighted average algorithm formula is:

[0051] in is the final estimated calculated value of the sun's position, , and They are the sun position data from machine vision, light intensity sensor, and temperature sensor. , and are the weights of the above sensor data respectively.

[0052] As an improvement to the above embodiment, in order to cope with environmental changes and fluctuations in data quality of different sensors, the weight distribution is determined by a dynamic weight adjustment mechanism: 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 502 can be judged by the change in the light intensity value. When the light intensity changes slightly 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 other sensor data. When its data matches well with other sensor data, its weight can be increased.

[0053] As an improvement of the above embodiment, the image processing algorithm includes Collect images of the target object, the sun, in advance; select different shooting angles and shooting time periods to collect image data, and filter the images to retain images with higher resolution; The collected image information is subjected to binary segmentation using the maximum inter-class variance method, thereby generating a binary image for subsequent processing; Annotate the sun in the acquired images with bounding boxes and assign a unique identifier to each marker. This process involves determining the length, width, and center coordinates of the bounding box to ensure that each sun target can be accurately identified and located in subsequent analysis. The image is selected and divided into equal parts using Python and OpenCV libraries to achieve quantitative analysis of the intensity and direction of sunlight. This step includes setting the height and width of the image matrix and determining the two vertex positions of the sun positioning area. Different areas are distinguished by color triplets, thereby visually ensuring clarity and improving recognition accuracy. Convert the processed images to PNG format to ensure that the images maintain high quality during storage and transmission, which is convenient for subsequent analysis and model training. The corresponding solar signal dataset is created using the images that have been annotated and drawn in fixed areas. It is divided into training set, test set, and validation set in proportion and passed into a single-stage anchor-free deep convolutional neural network. In this step, the optimal pre-trained YOLO model is obtained through forward propagation and parameter update of the closed-loop system to improve the accuracy of target detection and analysis. When working, the camera obtains video information of the sun target in the sky in real time, and the process reads the video stream frame by frame; The data is input into the optimal pre-trained model to achieve quantitative analysis of the center position of the sun in the sky and the range of solar light intensity in the anchor frame: the solar target is detected, its center position is located, and the range of light intensity in the anchor frame is determined to obtain the position information of the sun relative to the photovoltaic panel.

[0054] As an improvement of the above embodiment, the fixed area drawing method includes: Define an image as a The matrix of is the height of the image (number of rows), is the width of the image (number of columns); and They represent the two vertices of the fixed selection area of ​​the sun positioning center. To distinguish the sun center identification box from the fixed selection area, the color is defined as a triplet. , indicating blue ( ),green( ) and red ( ), the specific formula is as follows:

[0055]

[0056]

[0057]

[0058]

[0059] In the image matrix The operation of drawing a fixed selection area rectangle for quantitative analysis of solar radiation intensity can be expressed as:

[0060]

[0061] Then the selected area for the drawn solar radiation intensity analysis is divided into three equal parts and labeled as areas A, B and C; Indicates the bounding box thickness.

[0062] As an improvement of the above embodiment, the quantitative analysis of the range of sunlight intensity in the anchor frame includes: definition A quantitative index of the solar intensity in each third of the fixed selected area, It represents the overlapping area of ​​the sun center identification box and the fixed selection area in every one-third of the fixed area. represents the area of ​​a single one-third fixed selection area; the specific formula for quantitative analysis of the sunlight intensity in the anchor frame is:

[0063] Define three thresholds, Threshold 1, Threshold 2, and Threshold 3. When these thresholds are exceeded, the solar radiation intensity is judged to be mild, moderate, and severe; The quantitative analysis formula of light intensity is as follows:

[0064] in Represents the overlapping area of ​​the target area and the light source illumination area, Represents the total area of ​​the fixed selection region.

[0065] As an improvement of the above embodiment, 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 sunlight source, which can be achieved by calculating the deviation between the center point of the target and the sunlight source:

[0066]

[0067] In the formula , Indicates the center position coordinates of the target, , Indicates the current coordinates of the sun light source; In the process of light chasing, the target position and the position of the tracking light source are continuously adjusted to form a closed-loop control to ensure the stability of the light chasing effect. The adjustment formula in this process is:

[0068] in is the feedback coefficient to control the strength of the feedback.

[0069] 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 the Kalman filter. The specific formula is:

[0070]

[0071] In the formula is the target state, such as position, speed, etc.; is the state transfer matrix, To control the input model, is the control input, is the process noise, is the observed value, is the observation matrix, is the observation noise.

[0072] Based on this, this embodiment can realize the correction of the calculation deviation of the sun position coordinates. The sun position coordinates at the next moment predicted by Kalman filtering can be used as a reference or correction value to adjust the actual measurement value and the tracking strategy of the system. For example: Correct the current deviation: During real-time tracking, if there is a large deviation between the current measurement value (such as light intensity or visual data) and the Kalman filter prediction value, the deviation can be applied to correct the current coordinates to improve tracking accuracy.

[0073] Prediction error correction: The prediction results provided by the Kalman filter are used as a benchmark reference. If the real-time measured sun position deviates from the predicted value by more than a certain threshold, the system is triggered to adjust or recalibrate.

[0074] The predictive ability of Kalman filtering is used to "guide" the current tracking path, allowing the solar photovoltaic system to respond to changes in the sun's position in advance and reduce error accumulation due to sensor noise or environmental changes.

[0075] Example 3 Under the action of the real-time tracking and detection method of the solar signal, this embodiment provides a working method of a solar photovoltaic panel self-cleaning device, including: Step S1: start the height adaptive device 3, and adjust the height of the LED lamp 7 according to the height of the plant under the LED lamp 7 to better promote the transpiration of the plant; Step S2: Start the sensing system 5, and obtain the solar signal information in real time through the solar signal real-time tracking detection method. When the solar signal position information deviates from the center position of the photovoltaic panel 6, transmit the signal to the control system through the serial port protocol; Step S3: the automatic light-chasing mechanism 2 adjusts the angle of the photovoltaic panel 6 according to the control signal, so that the photovoltaic panel 6 always faces the solar signal vertically to achieve better power generation efficiency; Step S4: while the solar photovoltaic panel 6 is working, the cleaning system 4 transports the water collected by plant transpiration to the high-pressure fine mist nozzle 401 through the diaphragm pump 402 to complete the self-cleaning of the solar photovoltaic panel 6.

[0076] It should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

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

2. The photovoltaic system based on transpiration self-cleaning according to claim 1, characterized in that: When the height adaptive mechanism is raised, the light sources on both sides are retracted to reduce the probability of damage to the mechanism; when the height adaptive mechanism is lowered, the light sources on both sides are turned on, and the distance between the light sources on both sides becomes larger, thereby increasing the illuminated area. 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.

3. The photovoltaic system based on transpiration self-cleaning according to claim 1, characterized in that: A ratchet structure is provided in the anti-rotation yoke assembly of the highly adaptive mechanism, so that the motor is powered on only when it needs to work, and is not powered on at other times. The current position is maintained by the self-locking of the ratchet, thereby achieving the purpose of reducing power consumption.

4. The photovoltaic system based on transpiration self-cleaning according to claim 1, characterized in that: 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 solar signal. The real-time tracking and detection method of the solar signal includes: The camera captures the sun image in real time, extracts the position and morphological characteristics 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; The light intensity sensor monitors the changes in solar radiation intensity in real time, and calculates the sun's azimuth and altitude angle based on the light intensity data measured at different angles. At the same time, the temperature sensor 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. Data fusion and tracking algorithm: By fusing the data from machine vision, light intensity sensor and temperature sensor, 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.

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 collect the sun's position, radiation intensity, and temperature changes in real time to obtain three types of data: visual data , light intensity data , 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 weighted average calculation, calculate the weighted data and output the final calculated value of the sun's position.

6. The photovoltaic system based on transpiration self-cleaning according to claim 5, characterized in that: In order to cope with environmental changes and fluctuations in data quality from different sensors, a dynamic weight adjustment mechanism is used to determine weight distribution: 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 in the light intensity value. When the light intensity change is small 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 other sensor data. When its data matches well with other sensor data, 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-collecting an image of the target object, the sun; The collected image information is subjected to binary segmentation using the maximum inter-class variance method, thereby generating a binary image for subsequent processing; Annotate the sun in the acquired images with bounding boxes and assign a unique identifier to each marker. This process involves determining the length, width, and center coordinates of the bounding box to ensure that each sun target can be accurately identified and located in subsequent analysis. The image is selected and divided into equal parts using Python and OpenCV libraries to achieve quantitative analysis of the intensity and direction of sunlight. This step includes setting the height and width of the image matrix and determining the two vertex positions of the sun positioning area. Different areas are distinguished by color triplets, thereby visually ensuring clarity and improving recognition accuracy. Convert the processed images to PNG format to ensure that the images maintain high quality during storage and transmission, which is convenient for subsequent analysis and model training. The corresponding solar signal dataset is created using the images with completed annotations and fixed area drawing, and is passed into a single-stage anchor-free deep convolutional neural network. In this step, the optimal pre-training model is obtained through forward propagation and parameter update of the closed-loop system to improve the accuracy of target detection and analysis. When working, the camera obtains video information of the sun target in the sky in real time, and the process reads the video stream frame by frame; The data is input into the optimal pre-trained model to achieve quantitative analysis of the center position of the sun in the sky and the range of solar light intensity in the anchor frame: the solar target is detected, its center position is located, and the range of light intensity in the anchor frame is determined 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: Quantitative analysis of the range of sunlight intensity within the anchor frame includes definition A quantitative index of the solar intensity in each third of the fixed selected area, It represents the overlapping area of ​​the sun center identification box and the fixed selection area in every one-third of the fixed area. represents the area of ​​a single one-third fixed selection area; the specific formula for quantitative analysis of the sunlight intensity in the anchor frame is: Define three thresholds, Threshold 1, Threshold 2, and Threshold 3. When these thresholds are exceeded, the solar radiation intensity is judged to be mild, moderate, and severe, respectively.

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. This can be achieved by calculating the deviation between the center point of the target and the light source: In the formula , Indicates the center position coordinates of the target, , Indicates the current coordinates of the light source.

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

Images