Photovoltaic module cleaning method, device and equipment and computer readable storage medium

Abstract

The invention discloses a photovoltaic module cleaning method, device and equipment and a computer readable storage medium. The method comprises the steps that first collection data and second collection data of each subarea of the photovoltaic power station are acquired, the first collection data comprise illumination intensity and visible light images, the second collection data comprise n infrared images corresponding to n continuous inspection periods, and n is a positive integer larger than 1; for each sub-region, performing stain identification based on the first acquisition data to obtain a first identification result; performing hot spot region identification based on the second collection data to obtain a second identification result; and if the ratio of the stain area in the first identification result corresponding to the target sub-region is greater than a first threshold value or the second identification result is that the n infrared images all have hot spot regions, starting a spraying device corresponding to the target sub-region so as to clean the photovoltaic module in the target sub-region. The precision and economical efficiency of cleaning are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of photovoltaic system operation and maintenance technology, specifically to a method, apparatus, equipment, and computer-readable storage medium for cleaning photovoltaic modules. Background Technology

[0002] Existing automatic sprinkler systems lack the ability to sense the actual condition of photovoltaic modules, and most adopt a mode of starting at fixed time intervals. In this mode, when there is only a small amount of dust on the surface of the photovoltaic modules, the device may still spray the entire area according to the preset program, resulting in a waste of water and energy; while when some areas are severely stained and affect power generation efficiency, the device may not be able to clean them in time because the set time has not yet arrived, resulting in power generation loss. Summary of the Invention

[0003] To address the aforementioned technical problems, this application provides a method, apparatus, device, and computer-readable storage medium for cleaning photovoltaic modules.

[0004] In a first aspect, embodiments of this application provide a method for cleaning photovoltaic modules, the method comprising: Acquire first and second data for each sub-region of the photovoltaic power station. The first data includes light intensity and visible light images, and the second data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. For each sub-region, stain identification is performed based on the first collected data to obtain a first identification result; hot spot region identification is performed based on the second collected data to obtain a second identification result. If the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than the first threshold, or if the second identification result is that all n infrared images contain hot spot areas, then the spray device corresponding to the target sub-region is turned on to clean the photovoltaic modules in the target sub-region.

[0005] In conjunction with the first aspect, in one embodiment, the photovoltaic module cleaning method further includes: Based on the stain type in the first identification result corresponding to the target sub-region, set the spraying mode and spraying duration of the spraying device corresponding to the target sub-region.

[0006] In conjunction with the first aspect, in one implementation, the step of identifying stains based on the first collected data to obtain a first identification result includes: The first collected data is input into the stain recognition model to obtain the first recognition result. The stain recognition model is obtained by training YOLOv8.

[0007] In conjunction with the first aspect, in one implementation, before performing stain identification based on the first collected data for each sub-region to obtain a first identification result; and before performing hot spot region identification based on the second collected data to obtain a second identification result, the method further includes: Perform dehazing, noise reduction, and distortion correction on visible light and infrared images.

[0008] In conjunction with the first aspect, in one embodiment, before acquiring the first and second data collection data for each sub-region of the photovoltaic power station, the method further includes: During each inspection cycle, the drone is controlled to move along the inspection path. During the movement, the drone collects visible light images and infrared images of the photovoltaic power station at preset time intervals, and records the location and light intensity at the time of collection.

[0009] In conjunction with the first aspect, in one embodiment, after activating the spray device corresponding to the target sub-area to clean the photovoltaic modules in the target sub-area, the method further includes: Obtain new first data corresponding to the target sub-region; Stain identification is performed based on the new first collected data, resulting in a new first identification result. The stain residue rate is obtained based on the percentage of stain area in the new first identification result and the percentage of stain area in the first identification result. If the stain residue rate is less than the second threshold, the next inspection time for the target sub-area will be postponed by a preset time.

[0010] In conjunction with the first aspect, in one embodiment, obtaining the stain residue rate based on the stain area ratio in the new first identification result and the stain area ratio in the first identification result includes: Calculate the ratio of the percentage of stained area in the new first identification result to the percentage of stained area in the first identification result, and use the ratio as the stain residue rate.

[0011] Secondly, embodiments of this application provide a photovoltaic module cleaning device, the photovoltaic module cleaning device comprising: The acquisition module is used to acquire first and second acquisition data for each sub-area of ​​the photovoltaic power station. The first acquisition data includes light intensity and visible light images, and the second acquisition data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. The identification module is used to identify stains based on the first collected data for each sub-region, and obtain a first identification result; and to identify hot spot areas based on the second collected data, and obtain a second identification result. The spray control module is used to activate the spray device corresponding to the target sub-region if the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than a first threshold or the second identification result is that all n infrared images have hot spot areas, so as to clean the photovoltaic modules in the target sub-region.

[0012] Thirdly, embodiments of this application provide a photovoltaic module cleaning device, which includes a processor, a memory, and a photovoltaic module cleaning program stored in the memory and executable by the processor, wherein when the photovoltaic module cleaning program is executed by the processor, it implements the steps of the photovoltaic module cleaning method as described in the first aspect.

[0013] Fourthly, embodiments of this application provide a computer-readable storage medium storing a photovoltaic module cleaning program, wherein when the photovoltaic module cleaning program is executed by a processor, it implements the steps of the photovoltaic module cleaning method as described in the first aspect.

[0014] The beneficial effects of the technical solutions provided in this application include: In this embodiment, first and second acquisition data are obtained for each sub-region of the photovoltaic power station. The first acquisition data includes light intensity and visible light images, and the second acquisition data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. For each sub-region, stain identification is performed based on the first acquisition data to obtain a first identification result. Hot spot area identification is performed based on the second acquisition data to obtain a second identification result. If the proportion of stain area in the first identification result corresponding to the target sub-region is greater than a first threshold, or if the second identification result shows that hot spot areas exist in all n infrared images, then the spray device corresponding to the target sub-region is activated to clean the photovoltaic modules in the target sub-region. Through this embodiment, based on the dynamic perception of stain and hot spot status through multi-source data (visible light images, infrared images, and light intensity), on-demand cleaning is achieved, significantly improving the accuracy and economy of cleaning. Attached Figure Description

[0015] Figure 1 This is a schematic flowchart of an embodiment of the photovoltaic module cleaning method of this application; Figure 2 This is a functional module diagram of an embodiment of the photovoltaic module cleaning device of this application; Figure 3 This is a schematic diagram of the hardware structure of the photovoltaic module cleaning equipment involved in the embodiments of this application. Detailed Implementation

[0016] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0018] In a first aspect, embodiments of this application provide a method for cleaning photovoltaic modules.

[0019] In one embodiment, reference is made to Figure 1 , Figure 1 This is a schematic flowchart of an embodiment of the photovoltaic module cleaning method of this application. Figure 1 As shown, the photovoltaic module cleaning method includes: Step S10: Obtain first and second acquisition data for each sub-region of the photovoltaic power station. The first acquisition data includes light intensity and visible light images, and the second acquisition data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. In this embodiment, a large photovoltaic power station can be divided into multiple sub-regions (e.g., each sub-region corresponds to a set of photovoltaic module arrays). Each sub-region is equipped with a data acquisition system, including a visible light camera, an infrared camera, and a light sensor. The system performs inspections periodically (e.g., every morning): first, it acquires the first acquisition data (i.e., light intensity and visible light image) and the second acquisition data (i.e., 3 infrared images corresponding to n=3 consecutive inspection cycles) for each sub-region.

[0020] Alternatively, data can be collected from the entire photovoltaic power station using drones, and first and second collected data for each sub-area of ​​the photovoltaic power station can be obtained based on the data collected by the drones. Further, in one embodiment, before step S10, the following steps are also included: During each inspection cycle, the drone is controlled to move along the inspection path. During the movement, the drone collects visible light images and infrared images of the photovoltaic power station at preset time intervals, and records the location and light intensity at the time of collection.

[0021] In this embodiment, a drone (equipped with a visible light camera, an infrared camera, a light intensity sensor, and a GPS module) is controlled to fly along a preset inspection path. During each inspection cycle (e.g., once a day), the drone collects images at fixed time intervals (e.g., every 10 seconds) and simultaneously records its position coordinates and light intensity data, transmitting these data in real time to the execution entity in this embodiment. Upon receiving the data from the drone, the execution entity, for example, maps the position coordinates corresponding to the time the visible light image was collected to a photovoltaic module distribution map to determine which sub-region the visible light image corresponds to, thus determining the visible light image corresponding to each sub-region. Simultaneously, the light intensity corresponding to the time the visible light image was collected is associated with it to determine the visible light image and light intensity corresponding to each sub-region. Similarly, for infrared images collected during any inspection cycle, the position coordinates corresponding to the time the infrared image was collected are mapped to a photovoltaic module distribution map to determine which sub-region the infrared image corresponds to, thus determining the infrared image corresponding to each sub-region. This same processing is applied to the infrared images collected during each inspection cycle to determine the n infrared images collected over n consecutive inspection cycles corresponding to each sub-region. The value of n can be set according to actual needs, for example, it can be set to 2.

[0022] The improved A-Star algorithm is used to generate inspection paths. This algorithm combines the layout coordinates of the power station's photovoltaic array (preset in the system database) and introduces an "obstacle avoidance weight factor" on the basis of the traditional A-Star algorithm to avoid obstacles on the water surface (such as aquaculture cages and utility poles), ensuring that the UAV cruises at a constant speed along the rows of photovoltaic modules. The distance between adjacent inspection routes is set to 1.2 times the width of the photovoltaic panel to avoid missed inspections.

[0023] Step S20: For each sub-region, stain identification is performed based on the first collected data to obtain a first identification result; hot spot region identification is performed based on the second collected data to obtain a second identification result; In this embodiment, for each sub-region: based on visible light images, computer vision algorithms are used to identify stains (such as dust, bird droppings, etc.) and calculate the stain area percentage; based on infrared image sequences, hot spot regions are detected (indicating local overheating, possibly caused by stain occlusion). If the stain area percentage of a target sub-region is greater than a first threshold (e.g., 5%), or if all three infrared images show the presence of hot spot regions, the corresponding spray device (such as a high-pressure water gun or spray system) for that sub-region is automatically activated for targeted cleaning. Specifically, a U-Net-based semantic segmentation algorithm is used to process the infrared images and identify hot spot regions.

[0024] Furthermore, in one embodiment, the step of identifying stains based on the first collected data to obtain a first identification result includes: The first collected data is input into the stain recognition model to obtain the first recognition result. The stain recognition model is obtained by training YOLOv8.

[0025] In this embodiment, a stain recognition model is constructed based on the improved YOLOv8 algorithm. The model is trained by annotating photovoltaic panel images with different lighting conditions and different stain types (dust, bird droppings, water stains) to ensure that the model has a high accuracy in stain recognition and can output the stain area percentage.

[0026] Furthermore, in one embodiment, before step S10, the method further includes: Perform dehazing, noise reduction, and distortion correction on visible light and infrared images.

[0027] In this embodiment, the visible light image and infrared image are dehazed (using the dark channel prior algorithm), denoised (Gaussian filtering), and distortion corrected to eliminate the influence of water vapor and light reflection on image quality.

[0028] Step S30: If the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than the first threshold or the second identification result is that all n infrared images have hot spot areas, then the spray device corresponding to the target sub-region is turned on to clean the photovoltaic modules in the target sub-region.

[0029] Furthermore, in one embodiment, the photovoltaic module cleaning method further includes: Based on the stain type in the first identification result corresponding to the target sub-region, set the spraying mode and spraying duration of the spraying device corresponding to the target sub-region.

[0030] In this embodiment, the target sub-region is mapped to the pipeline coordinate system of the automatic sprinkler module to determine the sprinkler device that needs to be activated. Specifically, visible light images are used to identify stains as dust, bird droppings, or oil. Spray parameters are adjusted according to the stain type: for dust, the spray mode is set to low-pressure spray for 5 minutes; for bird droppings, the spray mode is set to high-pressure water flow for 10 minutes.

[0031] Furthermore, in one embodiment, after step S30, the method further includes: Step S40: Obtain new first collected data corresponding to the target sub-region; Step S50: Based on the new first collected data, stain identification is performed to obtain a new first identification result; Step S60: Obtain the stain residue rate based on the stain area ratio in the new first identification result and the stain area ratio in the first identification result; In step S70, if the stain residue rate is less than the second threshold, the next inspection time for the target sub-area will be postponed by a preset time.

[0032] In this embodiment, immediately after spray cleaning of the target sub-area, the drone is controlled to re-acquire visible light images of the area (new first acquisition data). The same stain recognition model is used to obtain a new stain area percentage, and then the stain residue rate is calculated. If the residue rate is less than a second threshold (e.g., 5%), the next inspection time for the sub-area is postponed from 24 hours to 48 hours.

[0033] Further, in one embodiment, step S60 includes: Calculate the ratio of the percentage of stained area in the new first identification result to the percentage of stained area in the first identification result, and use the ratio as the stain residue rate.

[0034] In this embodiment, assuming that the stain area accounts for 8% in the first identification result before cleaning and the stain area accounts for 0.8% in the new first identification result after cleaning, the stain residue rate is 10%.

[0035] The following section will further illustrate this with a specific scenario: (1) System architecture setup Drone inspection module: Equipped with a high-definition visible light camera, infrared thermal imager and Beidou positioning device, it is responsible for collecting surface images, temperature data and location information of photovoltaic modules. The drone is a highly reliable and protected industrial model, which is suitable for the aquatic environment of the fishery-solar hybrid power station.

[0036] Automatic sprinkler module: It consists of sprinklers distributed in the photovoltaic array, water supply network, water pump, etc. The sprinkler control unit can remotely control the area of ​​the sprinkler components, water pressure and spray duration through the data processing center.

[0037] Data processing center: Deployed in the power plant monitoring room or communication and relay protection room, it integrates edge computing servers and cloud platforms, and has the functions of real-time data reception, analysis and command issuance, and supports communication with drones and sprinkler equipment.

[0038] (2) Intelligent inspection and data collection by unmanned aerial vehicles Path planning: An improved A* algorithm is used to generate inspection paths. This algorithm combines the layout coordinates of the power station's photovoltaic array (preset in the system database) and introduces an "obstacle avoidance weight factor" on the basis of the traditional A* algorithm to avoid obstacles on the water surface (such as aquaculture cages and utility poles), ensuring that the UAV cruises at a constant speed along the rows of photovoltaic modules. The spacing between adjacent inspection routes is set to 1.2 times the width of the photovoltaic panel to avoid missed inspections.

[0039] Data Acquisition: When the UAV flies along the planned path, the visible light camera takes images of the photovoltaic module surface at fixed intervals, the infrared thermal imager collects temperature data simultaneously, and the Beidou positioning module records the real-time location. All data can be transmitted back in real time through the module or transmitted after returning to the UAV nest.

[0040] (3) Data processing and fault identification Image preprocessing: The data processing center performs dehazing (using dark channel prior algorithm), noise reduction (Gaussian filtering), and distortion correction on the returned images to eliminate the impact of water vapor and light reflection on image quality.

[0041] Stain and Fault Identification: A stain identification model is built based on the improved YOLOv8 algorithm. The model is trained by annotating photovoltaic panel images with different lighting and different stain types (dust, bird droppings, water stains) to ensure that the model has a high accuracy in identifying stains and can output the stain area percentage.

[0042] A semantic segmentation algorithm based on U-Net is used to process infrared thermal images to identify hot spot regions. At the same time, the specific location of the faulty component is located by combining the component circuit topology.

[0043] Cleaning demand determination: When the area of ​​stains accounts for ≥3% or the hot spot area persists (detected in 2 consecutive inspections), the system determines that the area needs to be sprayed for cleaning.

[0044] (4) Automatic sprinkler coordinated control Spraying area positioning: Based on the location and component layout coordinates transmitted back by the drone, the data processing center maps the area to be cleaned to the pipeline coordinate system of the automatic spraying module, generating a spraying coordinate list. Spray parameter optimization: Parameters are dynamically adjusted based on stain type. For dusty stains: low-pressure, long-duration spraying is used, combined with 360-degree full coverage from the nozzles. For stubborn stains (bird droppings, mud stains): the corresponding stain focusing mode of the spray device is activated, such as high-pressure or pulse spraying. The above parameters are adjusted in real time using a fuzzy PID control algorithm to ensure that the water pressure remains stable within the set range. Linked execution: The data processing center sends switching commands to the solenoid valves in the corresponding areas, and the sprinkler system performs cleaning operations according to optimized parameters. At the same time, the IoT module provides real-time feedback on the sprinkler status (on / off, water pressure, flow rate). (5) Closed-loop verification and dynamic adjustment Effect re-inspection: After spraying, the drone will conduct a second inspection of the cleaned area along the original path. The algorithm in step three will be used to obtain the new stain area ratio and evaluate the stain residue rate (residue rate = stain area after cleaning / stain area before cleaning).

[0045] Dynamic optimization: If the residue rate is ≤5%, the cleaning is deemed satisfactory, the system updates the cleaning record for that area, and the next inspection cycle is extended by 20% (adaptive adjustment based on historical data). If the residue rate is >5%, a second spray is automatically triggered, and the area is marked as a "key concern area" for priority coverage in the next inspection, while the water pressure parameters for the next spray are optimized. Global collaborative scheduling: The system can automatically generate maintenance reports periodically, including data such as cleaning coverage rate, fault handling rate, and power generation efficiency impact assessment. It dynamically plans the priority of inspection and cleaning tasks for the next day using the Particle Swarm Optimization (PSO) algorithm, achieving optimal allocation of resources (drone endurance, sprinkler water consumption). Through the above embodiments, this solution achieves a closed-loop collaborative process of "inspection-decision-cleaning-verification," solving problems such as the disconnect between cleaning and inspection, low efficiency, and high costs in existing technologies, significantly improving the intelligent operation and maintenance level of the solar-aquaculture hybrid power station.

[0046] Secondly, embodiments of this application also provide a photovoltaic module cleaning device.

[0047] In one embodiment, reference is made to Figure 2 , Figure 2 This is a functional module diagram of an embodiment of the photovoltaic module cleaning device of this application. Figure 2 As shown, the photovoltaic module cleaning device includes: The acquisition module 10 is used to acquire first acquisition data and second acquisition data for each sub-area of ​​the photovoltaic power station. The first acquisition data includes light intensity and visible light images, and the second acquisition data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. The identification module 20 is used to identify stains based on the first collected data for each sub-region, and obtain a first identification result; and to identify hot spot areas based on the second collected data, and obtain a second identification result. The spray control module 30 is used to activate the spray device corresponding to the target sub-region if the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than the first threshold or the second identification result is that all n infrared images have hot spot areas, so as to clean the photovoltaic modules in the target sub-region.

[0048] Furthermore, in one embodiment, the spray control module 30 is also used for: Based on the stain type in the first identification result corresponding to the target sub-region, set the spraying mode and spraying duration of the spraying device corresponding to the target sub-region.

[0049] Furthermore, in one embodiment, the identification module 20 is specifically used for: The first collected data is input into the stain recognition model to obtain the first recognition result. The stain recognition model is obtained by training YOLOv8.

[0050] Furthermore, in one embodiment, the photovoltaic module cleaning device further includes a pretreatment module for: Perform dehazing, noise reduction, and distortion correction on visible light and infrared images.

[0051] Furthermore, in one embodiment, the photovoltaic module cleaning device further includes a data collection module for: During each inspection cycle, the drone is controlled to move along the inspection path. During the movement, the drone collects visible light images and infrared images of the photovoltaic power station at preset time intervals, and records the location and light intensity at the time of collection.

[0052] Furthermore, in one embodiment, the photovoltaic module cleaning device further includes a cruise strategy adjustment module for: Obtain new first data corresponding to the target sub-region; Stain identification is performed based on the new first collected data, resulting in a new first identification result. The stain residue rate is obtained based on the percentage of stain area in the new first identification result and the percentage of stain area in the first identification result. If the stain residue rate is less than the second threshold, the next inspection time for the target sub-area will be postponed by a preset time.

[0053] Furthermore, in one embodiment, the cruise strategy adjustment module is specifically used for: Calculate the ratio of the percentage of stained area in the new first identification result to the percentage of stained area in the first identification result, and use the ratio as the stain residue rate.

[0054] The functions of each module in the photovoltaic module cleaning device correspond to the steps in the photovoltaic module cleaning method embodiment, and their functions and implementation processes will not be described in detail here.

[0055] Thirdly, embodiments of this application provide a photovoltaic module cleaning device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0056] Reference Figure 3 , Figure 3 This is a schematic diagram of the hardware structure of the photovoltaic module cleaning equipment involved in the embodiments of this application. In the embodiments of this application, the photovoltaic module cleaning equipment may include a processor, a memory, a communication interface, and a communication bus.

[0057] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0058] Communication interfaces include input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting devices within the photovoltaic module cleaning equipment, as well as interfaces used for interconnecting the photovoltaic module cleaning equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0059] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0060] The processor can be a general-purpose processor, which can call the photovoltaic module cleaning program stored in the memory and execute the photovoltaic module cleaning method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the photovoltaic module cleaning program is called can be referred to the various embodiments of the photovoltaic module cleaning method of this application, and will not be repeated here.

[0061] Those skilled in the art will understand that Figure 3 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0062] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0063] The present application provides a computer-readable storage medium storing a photovoltaic module cleaning program, wherein when the photovoltaic module cleaning program is executed by a processor, it implements the steps of the photovoltaic module cleaning method described above.

[0064] The method implemented when the photovoltaic module cleaning procedure is executed can be referred to in the various embodiments of the photovoltaic module cleaning method of this application, and will not be repeated here.

[0065] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0066] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0067] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0068] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0069] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0070] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0071] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for cleaning photovoltaic modules, characterized in that, The photovoltaic module cleaning method includes: Acquire first and second data for each sub-region of the photovoltaic power station. The first data includes light intensity and visible light images, and the second data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. For each sub-region, stain identification is performed based on the first collected data to obtain a first identification result; hot spot region identification is performed based on the second collected data to obtain a second identification result. If the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than the first threshold, or if the second identification result is that all n infrared images contain hot spot areas, then the spray device corresponding to the target sub-region is turned on to clean the photovoltaic modules in the target sub-region.

2. The photovoltaic module cleaning method as described in claim 1, characterized in that, The photovoltaic module cleaning method also includes: Based on the stain type in the first identification result corresponding to the target sub-region, set the spraying mode and spraying duration of the spraying device corresponding to the target sub-region.

3. The photovoltaic module cleaning method as described in claim 2, characterized in that, The stain identification based on the first collected data, to obtain the first identification result, includes: The first collected data is input into the stain recognition model to obtain the first recognition result. The stain recognition model is obtained by training YOLOv8.

4. The photovoltaic module cleaning method as described in claim 1, characterized in that, For each sub-region, stain identification is performed based on the first collected data to obtain a first identification result; Before obtaining the second identification result by identifying hot spot regions based on the second acquired data, the process also includes: Perform dehazing, noise reduction, and distortion correction on visible light and infrared images.

5. The photovoltaic module cleaning method as described in claim 1, characterized in that, Before acquiring the first and second data collection data for each sub-region of the photovoltaic power station, the method further includes: During each inspection cycle, the drone is controlled to move along the inspection path. During the movement, the drone collects visible light images and infrared images of the photovoltaic power station at preset time intervals, and records the location and light intensity at the time of collection.

6. The photovoltaic module cleaning method as described in claim 1, characterized in that, After activating the spray device corresponding to the target sub-area to clean the photovoltaic modules in the target sub-area, the method further includes: Obtain new first data corresponding to the target sub-region; Stain identification is performed based on the new first collected data, resulting in a new first identification result. The stain residue rate is obtained based on the percentage of stain area in the new first identification result and the percentage of stain area in the first identification result. If the stain residue rate is less than the second threshold, the next inspection time for the target sub-area will be postponed by a preset time.

7. The photovoltaic module cleaning method as described in claim 6, characterized in that, The process of obtaining the stain residue rate based on the stain area ratio in the new first identification result and the stain area ratio in the first identification result includes: Calculate the ratio of the percentage of stained area in the new first identification result to the percentage of stained area in the first identification result, and use the ratio as the stain residue rate.

8. A photovoltaic module cleaning device, characterized in that, The photovoltaic module cleaning device includes: The acquisition module is used to acquire first and second acquisition data for each sub-area of ​​the photovoltaic power station. The first acquisition data includes light intensity and visible light images, and the second acquisition data includes n infrared images corresponding to n consecutive inspection cycles, where n is a positive integer greater than 1. The identification module is used to identify stains based on the first collected data for each sub-region, and obtain a first identification result; and to identify hot spot areas based on the second collected data, and obtain a second identification result. The spray control module is used to activate the spray device corresponding to the target sub-region if the proportion of the stain area in the first identification result corresponding to the target sub-region is greater than a first threshold or the second identification result is that all n infrared images have hot spot areas, so as to clean the photovoltaic modules in the target sub-region.

9. A photovoltaic module cleaning device, characterized in that, The photovoltaic module cleaning device includes a processor, a memory, and a photovoltaic module cleaning program stored in the memory and executable by the processor, wherein when the photovoltaic module cleaning program is executed by the processor, it implements the steps of the photovoltaic module cleaning method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a photovoltaic module cleaning program, wherein when the photovoltaic module cleaning program is executed by a processor, it implements the steps of the photovoltaic module cleaning method as described in any one of claims 1 to 7.

Images