Cleaning device and method for cleaning solar cell panel in desert

Abstract

A cleaning device for cleaning a solar cell panel in a desert comprises a cleaning robot and a transport engineering vehicle for loading the cleaning robot, and the cleaning robot and the transport engineering vehicle form a primary-secondary system; the cleaning robot has the functions of autonomous driving and obstacle avoidance and is used for driving to the solar cell panel for cleaning operation, when the power of the cleaning robot is low, a low power signal is sent to the transportation engineering vehicle, and the transportation engineering vehicle autonomously drives to recover the cleaning robot for charging; the transportation engineering vehicle has autonomous navigation and obstacle avoidance functions and is used for transporting the cleaning robot to the position beside the solar cell panel according to a planned route, and the solar cell panel is installed at the top of the transportation engineering vehicle and used for supplying power to the transportation engineering vehicle and the cleaning robot; the transportation engineering vehicle adopts triangular crawler-type wheels. According to the cleaning device, efficient cleaning and flexible operation can be achieved, adaptability is high, stability is good, manpower can be reduced, and cost benefits and environment friendliness are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of desert solar cell panel cleaning, in particular to a cleaning device for cleaning solar cell panels in the desert, and also relates to a cleaning method of the above-mentioned cleaning device for cleaning solar cell panels in the desert. BACKGROUND

[0002] In recent years, with the vigorous development and application of new energy by the country, the solar photovoltaic power generation industry has developed rapidly, and the cleaning of solar cell panels has become a big problem, and the cleaning of solar cell panels has become increasingly urgent. According to statistics, the solar cell panel will reduce the power generation by 7%-35% due to the accumulation of dust on the surface without timely cleaning, thereby reducing the economic benefits of the entire power plant; manual cleaning in the mining area will consume a lot of manpower, material resources, financial resources and time.

[0003] Especially in the desert area, the wind and sand are large, and the sand dust is easy to accumulate on the surface of the solar cell panel, block the sunlight, reduce the photoelectric conversion efficiency of the cell panel, and even cause local temperature rise, resulting in component attenuation, backboard burning and other problems.

[0004] The existing solar cell panel cleaning technology mainly includes manual cleaning, water cleaning and automatic cleaning equipment, but manual cleaning has high safety risk, low efficiency and high cost; water cleaning is difficult to implement in the desert area due to lack of water resources; although the automatic cleaning equipment (such as track robot) can improve the cleaning efficiency, it has poor flexibility, insufficient adaptability and high maintenance cost. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a cleaning device for cleaning solar cell panels in the desert, which can realize efficient cleaning and flexible operation, has strong adaptability and good stability, and is helpful to reduce manpower, has cost-effectiveness and environmental protection.

[0006] Another technical problem to be solved by the present application is to provide a cleaning method of the above-mentioned cleaning device for cleaning solar cell panels in the desert.

[0007] The technical problem to be solved by the present application is solved by the following technical scheme. The present application is a cleaning device for cleaning solar cell panels in the desert, which comprises a cleaning robot and a transport engineering vehicle for loading the cleaning robot, and the cleaning robot and the transport engineering vehicle form a mother-daughter system;

[0008] The cleaning robot has autonomous driving and obstacle avoidance functions, and is used for driving to the solar cell panel to perform cleaning operation; when the cleaning robot has low power, a low power signal is sent to the transport engineering vehicle, and the transport engineering vehicle autonomously drives to the cleaning robot to recover and charge.

[0009] The transport engineering vehicle has autonomous navigation and obstacle avoidance functions, is used for transporting the cleaning robot to the side of the solar cell panel according to a planned route, and is provided with a solar cell panel on the top of the transport engineering vehicle and is used for supplying power to the transport engineering vehicle and the cleaning robot; the transport engineering vehicle adopts triangular track wheels, so that the contact area with the ground is increased and sinking into soft desert sand is prevented.

[0010] The technical problem to be solved by the present application can also be further realized by the following technical solutions.

[0011] A brush is used for cleaning dust on the solar cell panel and avoiding scratching the surface of the solar cell panel;

[0012] A water storage device with a high-pressure water spraying device is used for being connected with the transport engineering vehicle through a hose and spraying water for cleaning when needed;

[0013] A rubber track is used for preventing the cleaning robot from slipping on the solar cell panel;

[0014] An interface is used for being connected with the transport engineering vehicle and transmitting signals and water.

[0015] The technical problem to be solved by the present application can also be further realized by the following technical solutions.

[0016] The technical problem to be solved by the present application can also be further realized by the following technical solutions.

[0017] A high-definition camera is used for identifying the dust degree of the surface of the solar panel, judging whether water cleaning is needed, and serving as a sensing obstacle avoidance device;

[0018] A motor is used for driving the triangular track wheel to rotate;

[0019] A battery is used for storing the converted electric energy of the solar cell panel, providing electric energy for the transport engineering vehicle, and charging the cleaning robot;

[0020] A lifting platform is used for sending the cleaning robot to the solar cell panel;

[0021] A water tank is used for providing cleaning water for the cleaning robot;

[0022] Water pipes and signal transmission cables are used to connect with the cleaning robot to transmit signals and supply water.

[0023] The technical problem solved by the present application can also be further realized by the following technical solutions. For the cleaning device for cleaning solar panels in the desert described above, the solar panels mounted on the top of the transport engineering vehicle are detachable solar panels.

[0024] The technical problem solved by the present application can also be further realized by the following technical solutions. For the cleaning device for cleaning solar panels in the desert described above, the device uses an image processing algorithm to quantify the dust coverage area or concentration on the surface of the solar panel through gray scale analysis, edge detection, and color feature analysis, and automatically adjusts the water supply or cleaning method according to the analysis results.

[0025] The technical problem solved by the present application can also be further realized by the following technical solutions. For the cleaning device for cleaning solar panels in the desert described above, the cleaning device for cleaning solar panels in the desert has the following steps:

[0026] (1) Task planning and deployment

[0027] Set the cleaning route: the staff sets the planned route of the transport engineering vehicle through the control system, and the transport engineering vehicle has the functions of autonomous navigation and obstacle avoidance, and drives to the designated solar panel area according to the preset route;

[0028] Transport the cleaning robot: after the transport engineering vehicle arrives at the designated location, the lifting platform is used to lift multiple cleaning robots onto the solar panel, and each cleaning robot has the functions of autonomous driving and obstacle avoidance, and can independently carry out cleaning work;

[0029] (2) Panel surface detection and cleaning strategy development

[0030] Image acquisition and dirt analysis:

[0031] High-definition industrial cameras are used to collect images of the panel surface, and the high-definition industrial cameras are installed 1.5 m above the solar panel array, equipped with anti-glare light and dustproof and waterproof cover;

[0032] Image features are extracted through image processing algorithms, including gray scale standard deviation and dark pixel ratio, to determine the dirt level, including light pollution, moderate pollution and heavy pollution;

[0033] According to the dirt level, the cleaning method and key execution parameters are determined;

[0034] (3) Cleaning execution

[0035] Light pollution cleaning:

[0036] Adopt high pressure jet with brush cleaning method, first jet 10 seconds, blow away floating dust, then brush roll 20 seconds, clean attached dust, no water consumption in the whole process;

[0037] Moderate pollution cleaning:

[0038] Adopt high pressure jet, brush and micro mist spray combination method, first jet 10 seconds, then brush roll 20 seconds, finally intermittent mist spray 15 seconds, accurate wet stubborn dust;

[0039] Severe pollution cleaning:

[0040] Adopt low pressure water flow flushing, brush and jet drying combination method, first low pressure water flow flushing 15 seconds, flush off oil stains or bird droppings, then brush roll 20 seconds, finally jet 10 seconds, avoid water accumulation affecting power generation;

[0041] (4) Effect feedback and secondary adjustment

[0042] Cleaning effect evaluation:

[0043] After cleaning, the camera collects images again, and analyzes the dirt residue rate;

[0044] If the residue rate is less than 5%, it is determined to be up to standard, and the next panel cleaning is entered;

[0045] If the residue rate is between 5% and 10%, adjust the parameters and repeat the cleaning once;

[0046] If the residue rate is higher than 10%, trigger an alarm and notify manual inspection.

[0047] The technical problems to be solved by the present application can also be further realized by the following technical solutions, for the above-mentioned cleaning method for desert cleaning solar cell panel, the method further comprises

[0048] (1) Dynamic optimization of cleaning path and frequency:

[0049] Optimization based on sand accumulation prediction model:

[0050] Use historical data and real-time sensor data to build a solar panel surface sand accumulation prediction model;

[0051] AI dynamically optimizes the cleaning path and frequency according to the sand accumulation model, weather warning and battery status;

[0052] (2) Cleaning frequency adjustment strategy:

[0053] Emergency cleaning: immediately start cleaning when the weather warning is yellow or above, the real-time sand layer thickness is ≥0.8mm, or the power generation efficiency decreases by ≥15%; the robot runs at full load, and the whole station cleaning is completed within 2 hours;

[0054] Regular cleaning: when the sand layer thickness is 0.5-0.8mm, the light sand warning or the power generation efficiency decreases by 8%-15%, the cleaning frequency is adjusted to 2 days once, and the area with sand layer thickness ≥0.6mm is preferentially covered;

[0055] Low-frequency cleaning: when the sand layer thickness is <0.5mm, there is no sand warning or the power generation efficiency decreases by <8%, the cleaning frequency is extended to 7 days once, and only when the cleaning robot battery power is ≥80% is it executed.

[0056] Compared with the prior art, the beneficial effects of the present application are:

[0057] 1. Efficient cleaning and flexible operation

[0058] A transport engineering vehicle can load multiple cleaning robots to realize multi-machine cooperative operation, significantly improve cleaning efficiency, and is especially suitable for the rapid cleaning needs of large-area solar power stations;

[0059] 2. Trackless autonomous navigation

[0060] Break free from the limitations of traditional track-type robots, both the transport engineering vehicle and the cleaning robot have autonomous navigation and obstacle avoidance functions, and can flexibly adapt to complex desert terrain (such as sand dunes, deep sand areas) and irregularly arranged panel arrays;

[0061] 3. Anti-skid and anti-sinking design

[0062] The transport engineering vehicle is equipped with four triangular track-type wheels, which increase the contact area with the sand and improve the anti-skid and climbing ability, avoiding damage to the equipment caused by slipping or sinking into the sand;

[0063] 4. Intelligent cleaning strategy and water and energy saving

[0064] Quantify the dirtiness of the panel through image processing algorithms (such as gray scale analysis, edge detection), and automatically adjust the water supply or cleaning method (such as high-pressure air jet, micro-fog spraying, low-pressure water flow) through PID control to realize "cleaning on demand" and save water resources;

[0065] 5. Hybrid energy management

[0066] The transport engineering vehicle is equipped with solar panels on the top to power itself and the cleaning robot, and is also equipped with a battery energy storage system to extend the endurance time and reduce the dependence on external power supply; the intelligent energy management system dynamically allocates energy according to task requirements to improve utilization efficiency;

[0067] 6. AI path optimization

[0068] Based on historical data and real-time sensor information (such as sand layer thickness, weather warning), build a sand accumulation prediction model, dynamically adjust the cleaning path and frequency (such as cleaning in advance of sandstorm), improve efficiency and prolong the service life of the equipment;

[0069] 7、Modular design and maintainability

[0070] Cleaning robot and transport engineering vehicle realize data interaction (such as power state, cleaning task) through wireless signal transmission module, which is convenient for centralized management and scheduling; the robot can automatically return to the engineering vehicle for charging when the battery is depleted, reducing manual intervention;

[0071] 8、Cost effectiveness and environmental protection

[0072] Completely autonomous operation, no need for manual intervention, reducing labor costs and safety risks; water-free dry sweeping as the main, water cleaning as the auxiliary design, suitable for desert water shortage environment, avoiding water resource waste and transportation cost of traditional water washing method;

[0073] 9、Innovative cleaning brush head design

[0074] The cleaning robot is equipped with a new type of cleaning brush head, which combines high-pressure airflow or vibration technology to enhance cleaning strength while avoiding scratching the surface of the battery plate; the baffle design prevents water or dust from splashing, improving cleaning efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 is a structural schematic view of the present application;

[0076] Figure 2 is a structural front view of the present application;

[0077] Figure 3 is a structural side view of the present application;

[0078] Figure 4 is a structural schematic view of the cleaning robot of the present application;

[0079] Figure 5 is a structural side view of the cleaning robot of the present application. DETAILED DESCRIPTION

[0080] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0081] Reference Figures 1-5 A cleaning robot 7 device for desert cleaning solar panels adopts a parent-child system, a transport engineering vehicle is loaded with multiple cleaning robots 7, the transport engineering vehicle has autonomous navigation and obstacle avoidance functions, after the staff sets the planned route in advance, the transport engineering vehicle drives to the designated location according to the instructions, after arriving, the cleaning robot 7 is jacked up above the solar panel by the telescopic support, the cleaning robot 7 also has autonomous driving and obstacle avoidance capabilities, and then drives to the solar panel to carry out cleaning work;

[0082] When the cleaning robot 7 has low power, it automatically sends a signal to the transport engineering vehicle, the transport engineering vehicle autonomously drives to the cleaning robot 7 to recover it and charge it, the solar panel is installed on the top of the transport engineering vehicle for power supply for itself and the cleaning robot 7, and the water tank is provided on the vehicle to provide cleaning water for the cleaning robot 7, the transport engineering vehicle adopts four triangular track-type wheels to increase the contact area with the ground to prevent sinking in soft sand particles in the desert, especially in sand dunes and deep sand areas, and the anti-skid property and climbing ability of the track can avoid damage of the equipment due to skidding when moving irregularly between the panel arrays or working beside the inclined panel support.

[0083] I. Structure and function realization of cleaning robot 7

[0084] Driving mechanism: the cleaning robot 7 drives the rubber track 72 to move by rotating the gear 71, which can effectively prevent slipping on the solar panel to ensure stable movement;

[0085] Cleaning mechanism: a new type of cleaning brush head is provided, the brush 76 is driven to rotate by the rotating wheel 75 to clean the dust on the surface of the solar panel, the motor 77 provides power for the rotating wheel, and a high-pressure water spraying device 79 is provided, which is connected with the water storage device 73, when the panel is dusty, water is delivered to the high-pressure water spraying device through the hose for water spraying cleaning, the front support structure 74 plays a role in stable support, and the baffle 78 can prevent water and dust from flying around during cleaning to improve cleaning efficiency;

[0086] Signal and water supply interface: the back of the cleaning robot 7 is provided with an interface 80 connected with the transport engineering vehicle through the water supply pipe and the signal transmission cable to realize signal transmission and water supply.

[0087] II. Structure and function realization of transport engineering vehicle

[0088] Navigation and driving: the high-definition resolution camera 1 is used as a sensing and obstacle avoidance device, and the dust degree on the surface of the solar panel is identified through image processing technology to determine whether water cleaning is needed; the motor 2 drives the triangular track-type wheel 4 to rotate to realize the driving of the transport engineering vehicle, and the triangular track-type wheel design makes it not easy to sink into the sand particles;

[0089] Energy supply: A flip-up solar panel 9 is installed on the top to absorb light energy, convert it into electrical energy and store it in the battery 3, providing power support for the transport engineering vehicle itself and the cleaning robot 7;

[0090] Loading and transporting the cleaning robot 7: When the transport vehicle arrives at the designated location, the lifting platform 5 delivers the cleaning robot 7 to the solar panel, and the motor drives the lifting platform 5 to rise and fall. The transport vehicle can carry three cleaning robots 7, improving cleaning efficiency;

[0091] Water supply system: The water tank on the transport vehicle supplies water to the cleaning robot 7 through water pipes. Multiple water storage tanks are set up in the power station to supply water by gravity using the terrain elevation difference, reducing pump energy consumption. A ring network is formed by PE hoses to improve the reliability of water supply. Pressure sensors and solenoid valves are embedded in the network to monitor leaks in real time and automatically shut off faulty sections.

[0092] III. Implementation of Intelligent Cleaning Control System

[0093] Image acquisition and dirt analysis:

[0094] A high-definition industrial camera (1920×1080 resolution) is installed 1.5m above the solar panel array, equipped with an anti-glare supplementary light. The camera's field of view covers a single solar panel (approximately 1.6m×1m). Outdoors, a dustproof and waterproof cover is added. The camera acquires images at a preset frequency (3 images are taken per panel, and the clear, non-reflective image is selected) and transmits them via 4G / Ethernet to a PLC (such as a Siemens S7-1200) equipped with a PID controller. The image processing algorithm (such as OpenCV) automatically calculates the proportion of dark pixels and the grayscale standard deviation, compares it with a pre-stored threshold, and determines the level of contamination (light contamination: dark pixel proportion 5%-15%, grayscale standard deviation > 20; moderate contamination: dark pixel proportion 15%-30%, grayscale standard deviation > 40; heavy contamination: other situations).

[0095] PID control decision: The control objective is to achieve a dark pixel ratio of <5% and a grayscale standard deviation of <20 on the cleaned solar panel; the PLC calls the corresponding PID parameters (proportional coefficient P, integral time I, derivative time D) according to the level of dirtiness, as shown in the example below:

[0096]

[0097] Cleaning execution: PLC sends instructions to the execution end, and the electromagnetic valve adjusts the opening degree according to the PID parameters; when the pollution is light, first blow away the floating dust for 10s, and then roll the brush for 20s to clean the attached dust, without water consumption throughout the process; when the pollution is moderate, on the basis of air blowing and brush rolling, the micro-mist nozzle intermittently sprays (sprays for 5s and stops for 2s, for a total of 15s) to precisely moisten stubborn dust; when the pollution is heavy, first flush away the oil stains / bird droppings with low-pressure water flow for 15s, then roll the brush for 20s, and finally blow air for 10s to avoid water accumulation affecting power generation.

[0098] Effect feedback and secondary regulation: After cleaning, the camera collects images again to analyze the residual rate of dirt. If the residual rate is <5% (up to standard), proceed to the next panel cleaning; if the residual rate is 5%-10% (not up to standard), the PID controller automatically adjusts the parameters (such as increasing the micro-mist flow to 10L / h when the pollution is moderate) and repeats the cleaning once; if the residual rate is >10% (severe non-standard), trigger an alarm and notify manual inspection (there may be stubborn stains or equipment failure).

[0099] Later maintenance: After daily cleaning, the PLC automatically records the dirt level of each battery panel, the water consumption of cleaning, and the PID adjustment data to generate a "cleaning efficiency report"; check the air valve filter (to prevent blockage) and brush wear (replace if worn more than 3mm) every week, and calibrate the camera white balance (to avoid misjudgment) every month; during the rainy season, reduce the cleaning frequency (once every 3 days), and automatically trigger emergency cleaning after a sandstorm (give priority to heavy pollution panels).

[0100] Four, sand dust interference resistance and dynamic optimization system implementation

[0101] Sand dust interference resistance sensor application:

[0102] Combine short-range high-precision millimeter wave radar, anti-fouling optical sensors (specific wavelength), ultrasonic sensors, inertial navigation units, and other sand dust interference resistance sensors to reduce the failure risk of traditional vision in a sand-diffused environment.

[0103] Sand dust accumulation prediction model and dynamic optimization:

[0104] Based on historical data and real-time sensor data (such as pressure / distance sensors to detect sand layer thickness), a sand dust accumulation prediction model for the surface of the battery panel is constructed. AI dynamically optimizes the cleaning path and frequency based on the sand dust accumulation model, weather warnings, and battery status, for example, focusing on cleaning before or just after a sandstorm, and reducing the cleaning frequency in lightly polluted areas to save energy. The specific cleaning frequency optimization strategy is as follows:

[0105]

[0106]

[0107] Six, dust detection alternative method implementation

[0108] In addition to the above method of image acquisition and analysis by high-definition industrial cameras, the battery panel surface image can also be captured by a camera (industrial camera, unmanned aerial vehicle-mounted camera or mobile phone camera), and the image features (such as gray value, contrast, texture change) are analyzed by image processing algorithm to quantify the dust coverage area or concentration. The specific method includes:

[0109] Gray scale analysis: the gray scale value (brightness) of the dust-covered area image is lower than that of the clean area. The degree of dust accumulation is determined by calculating the "gray scale difference" or "low gray scale area proportion" (for example, set threshold: gray scale lower than 50 is "severe dust accumulation", 50-80 is "moderate", and above 80 is "mild");

[0110] Edge detection: dust particles cause image edge blur. The blur degree is calculated by edge sharpness algorithm (such as Canny operator). The higher the blur degree, the more serious the dust accumulation.

[0111] Color feature analysis: for color images, the color deviation of RGB channel is extracted (dust is usually gray / brown, which will change the original blue / black tone of the battery panel). This method has the advantages of being intuitive, non-contact and capable of large-area detection (such as unmanned aerial vehicle inspection), and can be automatically analyzed combined with AI algorithm. The disadvantages are that it is affected by light conditions (overcast, backlight affects image quality), battery panel surface reflection, shadow interference, and has low recognition accuracy for small and uniform dust (difficult to distinguish with the naked eye), and is suitable for regular inspection of large-scale photovoltaic power stations (unmanned aerial vehicle aerial photography + ground image comparison), portable detection of distributed power stations (handheld device photographing analysis).

Claims

1. A cleaning device for clearing solar panels in deserts, characterized in that: It includes a cleaning robot and a transport vehicle for loading the cleaning robot, forming a mother-daughter system; The cleaning robot has autonomous driving and obstacle avoidance functions. It is used to drive to the solar panels to carry out cleaning operations. When the cleaning robot's battery is low, it sends a low battery signal to the transport vehicle. The transport vehicle then autonomously drives over to retrieve the cleaning robot for charging. The transport vehicle has autonomous navigation and obstacle avoidance functions, and is used to transport the cleaning robot to the side of the solar panel according to the planned route. The solar panel is installed on the top of the transport vehicle to power the transport vehicle and the cleaning robot. The transport vehicle uses triangular tracked wheels to increase the contact area with the ground and prevent it from getting stuck in the soft sand of the desert.

2. The cleaning device for clearing solar panels in deserts according to claim 1, characterized in that: The cleaning robot includes: A brush is used to clean dust off solar panels while avoiding scratching the panel surface. A water storage device with a high-pressure water spraying device is used to connect to a transport vehicle via a hose and spray water for cleaning when needed; Rubber tracks are used to prevent the cleaning robot from slipping on the solar panels; The interface is used to connect with transport vehicles for signal and water transmission.

3. The cleaning device for clearing solar panels in deserts according to claim 2, characterized in that: The brush is mounted on the cleaning robot via a rotating wheel and a motor. A baffle is also installed on the cleaning robot at the brush mounting point to prevent water and dust from splashing everywhere.

4. The cleaning device for clearing solar panels in deserts according to claim 1, characterized in that: The transport engineering vehicle is also equipped with: A high-resolution camera is used to identify the level of dust on the surface of solar panels, determine whether water cleaning is needed, and also serves as a sensing obstacle avoidance device. A motor is used to drive the triangular track wheels to rotate. Batteries are used to store electrical energy converted from solar panels, providing power to transport vehicles and also charging cleaning robots. A lifting platform is used to deliver the cleaning robot to the solar panels; Water tank, used to provide cleaning water for the cleaning robot; Water pipes and signal cables are used to connect with the cleaning robot for signal transmission and water supply between the two.

5. The cleaning device for clearing solar panels in deserts according to claim 1 or 4, characterized in that: The solar panels mounted on the top of the transport vehicle are hinged solar panels.

6. The cleaning device for clearing solar panels in deserts according to claim 1, characterized in that: The device uses image processing algorithms to quantify the dust coverage area or concentration on the surface of solar panels through grayscale analysis, edge detection, and color feature analysis, and automatically adjusts the water supply or cleaning method based on the analysis results.

7. A method for cleaning solar panels used in desert dredging, characterized in that: This method uses the cleaning apparatus for clearing solar panels in the desert as described in any one of claims 1-6, and its steps are as follows: (1) Task planning and deployment Setting the cleaning route: Staff set the planned route for the transport vehicle through the control system. The transport vehicle has autonomous navigation and obstacle avoidance functions and travels to the designated solar panel area according to the preset route. Transport cleaning robots: After the transport engineering vehicle arrives at the designated location, it uses a lifting platform to lift multiple cleaning robots onto the solar panels. Each cleaning robot has autonomous driving and obstacle avoidance functions and can carry out cleaning operations independently. (2) Development of solar panel surface inspection and cleaning strategies Image acquisition and dirt analysis: High-definition industrial cameras are used to capture images of the surface of solar panels. The high-definition industrial cameras are installed 1.5m above the solar panel array and are equipped with anti-glare supplementary lights and dustproof and waterproof covers. Image features, including grayscale standard deviation and dark pixel ratio, are extracted using image processing algorithms to determine the level of contamination, including light contamination, moderate contamination, and heavy contamination. Based on the level of soiling, determine the cleaning method and key execution parameters; (3) Cleaning implementation Cleaning for light pollution: It uses a combination of high-pressure air jet and brush cleaning. First, it sprays air for 10 seconds to blow away the floating dust; then the brush rolls for 20 seconds to clean the attached dust. The whole process does not consume water. Cleaning for moderate pollution: It uses a combination of high-pressure jet spray, brush and micro-mist spray, first jet spray for 10 seconds, then brush roll for 20 seconds, and finally micro-mist nozzles spray intermittently for 15 seconds to precisely moisten stubborn dust. Heavy pollution cleaning: The process combines low-pressure water flushing, brushing, and air jet drying. First, low-pressure water flushes for 15 seconds to remove oil or bird droppings; then the brush rolls for 20 seconds; finally, air jets are used for 10 seconds to prevent water accumulation from affecting power generation. (4) Effect feedback and secondary adjustment Cleaning effectiveness evaluation: After cleaning, the camera captures images again to analyze the rate of dirt residue. If the residual rate is less than 5%, it is considered to meet the standard and proceeds to the next board for cleaning; If the residual rate is between 5% and 10%, fine-tune the parameters and repeat the cleaning process once. If the residual rate is higher than 10%, an alarm will be triggered, and a manual inspection will be notified.

8. The cleaning method for solar panels used in desert cleanup according to claim 7, characterized in that: The method also includes (1) Dynamically optimize cleaning path and frequency: Optimization based on the dust deposition prediction model: A predictive model for dust accumulation on the surface of solar panels was constructed using historical data and real-time sensor data. AI dynamically optimizes cleaning paths and frequencies based on sand and dust accumulation models, weather warnings, and battery status; (2) Cleaning frequency adjustment strategy: Emergency cleaning: When a yellow or higher weather warning for sandstorms is issued, the real-time sand layer thickness is ≥0.8mm, or the power generation efficiency drops by ≥15%, cleaning will be initiated immediately, with robots operating at full capacity to complete the cleaning of the entire station within 2 hours; Routine cleaning: When the sand layer thickness is 0.5-0.8mm, a light sandstorm warning is issued, or the power generation efficiency decreases by 8%-15%, the cleaning frequency is adjusted to once every 2 days, with priority given to covering areas with a sand layer thickness ≥0.6mm. Low-frequency cleaning: When the sand layer thickness is <0.5mm, there is no sandstorm warning, or the power generation efficiency drops by <8%, the cleaning frequency is extended to once every 7 days, and it is only performed when the cleaning robot's battery power is ≥80%.

Images