Photovoltaic module detection method, device, equipment and storage medium

Abstract

The application belongs to the technical field of unmanned aerial vehicles, and discloses a photovoltaic module detection method, device, equipment and storage medium. The method comprises the following steps: acquiring infrared image information according to an infrared camera; determining a target photovoltaic module image according to the infrared image information; and performing target detection on the target photovoltaic module image to determine a target hot spot. In the manner, the unmanned aerial vehicle is used for inspection, and the infrared imaging technology and the target detection technology are combined to automatically identify the hot spot of the photovoltaic module, so that the operation and maintenance efficiency of the photovoltaic equipment is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicles, and in particular to a photovoltaic module detection method, device, equipment and storage medium. BACKGROUND

[0002] At present, photovoltaic power station operation and maintenance mainly relies on the voltage and current and other electrical characteristics of photovoltaic inverters, but is limited by the installation mode of inverters and combiner boxes, and electrical operation and maintenance can only be accurate to the module string, and it is difficult to be accurate to specific modules, and is greatly affected by weather, so the fault diagnosis accuracy is not high. The operation and maintenance of specific photovoltaic modules mainly rely on manual inspection, and large photovoltaic power stations are distributed in complex environments, cover a large area, are affected by terrain, and are disordered and dispersed, so the manual inspection method is very time-consuming and laborious. For example, mountain power stations and water power stations bring great challenges to manual inspection.

[0003] The above content is only used to assist in understanding the technical solutions of the present application, and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0004] The main purpose of the present application is to provide a photovoltaic module detection method, device, equipment and storage medium, which aims to solve the technical problem of high difficulty of existing photovoltaic module inspection.

[0005] To achieve the above purpose, the present application provides a photovoltaic module detection method, which comprises the following steps:

[0006] According to the infrared camera, infrared image information is obtained;

[0007] According to the infrared image information, a target photovoltaic module image is determined;

[0008] The target photovoltaic module image is subjected to target detection to determine a target hot spot.

[0009] Optionally, the target photovoltaic module is subjected to target detection to determine a target hot spot, comprising:

[0010] The target photovoltaic module image is input into a preset target detection model to obtain a target detection result;

[0011] According to the target detection result, a target hot spot is determined.

[0012] Optionally, before the target photovoltaic module image is input into the preset target detection model to obtain the target detection result, it further comprises:

[0013] Obtain training samples, the training samples at least including normal photovoltaic module image samples, hot spot photovoltaic module image samples and reflective photovoltaic module image samples;

[0014] Train a preset initial neural network model according to the training sample, and obtain a preset target detection model.

[0015] Optionally, after the target photovoltaic module image is subjected to target detection to determine the target hot spot, the method further includes:

[0016] Obtaining unmanned aerial vehicle positioning information;

[0017] Monitoring relative position information of the target hot spot;

[0018] Determining a geographic position of the target photovoltaic module according to the unmanned aerial vehicle positioning information and the relative position information.

[0019] Optionally, after the target photovoltaic module image is subjected to target detection to determine the target hot spot, the method further includes:

[0020] Determining a pixel area of the target hot spot;

[0021] Determining a pixel area of a photovoltaic module corresponding to the target hot spot;

[0022] Determining an actual area of the target hot spot according to the pixel area of the target hot spot, the pixel area of the photovoltaic module corresponding to the target hot spot, and preset photovoltaic module product information, and determining a fault condition of the corresponding photovoltaic module according to the actual area of the target hot spot.

[0023] Optionally, after the infrared image information is obtained by the infrared camera, the method further includes:

[0024] Detecting image quality of the infrared image;

[0025] Generating a flight adjustment instruction value when the image quality of the infrared image is lower than preset image quality;

[0026] Adjusting flight parameters of the unmanned aerial vehicle according to the flight adjustment instruction until the image quality is no longer lower than the preset image quality.

[0027] Optionally, adjusting the flight parameters of the unmanned aerial vehicle according to the flight adjustment instruction until the image quality is no longer lower than the preset image quality includes:

[0028] Adjusting a flight height of the unmanned aerial vehicle according to the flight adjustment instruction;

[0029] Adjusting a flight speed of the unmanned aerial vehicle until the image quality is no longer lower than the preset image quality when the flight height adjustment of the unmanned aerial vehicle is completed and the image quality is lower than the preset image quality.

[0030] In addition, to achieve the above object, the application further provides a photovoltaic module detection device, which comprises:

[0031] an acquisition module configured to acquire infrared image information according to the infrared camera;

[0032] a processing module configured to determine a target photovoltaic component image according to the infrared image information;

[0033] The processing module is further configured to perform target detection on the target photovoltaic component image to determine a target hot spot.

[0034] In addition, to achieve the above object, the present application further provides a photovoltaic component detection device, which comprises a memory, a processor and a photovoltaic component detection program stored in the memory and executable on the processor, and the photovoltaic component detection program is configured to implement the steps of the photovoltaic component detection method as described above.

[0035] In addition, to achieve the above object, the present application further provides a storage medium, which stores a photovoltaic component detection program, and the photovoltaic component detection program implements the steps of the photovoltaic component detection method as described above when executed by a processor.

[0036] The present application acquires infrared image information according to the infrared camera, determines a target photovoltaic component image according to the infrared image information, and performs target detection on the target photovoltaic component image to determine a target hot spot. In this way, the photovoltaic component hot spot is automatically identified by using a UAV for inspection and combining infrared imaging technology and target detection technology, thereby improving the operation and maintenance efficiency of the photovoltaic device. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a structural schematic diagram of a photovoltaic component detection device of a hardware running environment related to the embodiment scheme of the present application;

[0038] Figure 2 is a flowchart of a first embodiment of the photovoltaic component detection method of the present application;

[0039] Figure 3 is a flowchart of a second embodiment of the photovoltaic component detection method of the present application;

[0040] Figure 4 is a structural block diagram of a first embodiment of the photovoltaic component detection device of the present application.

[0041] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0042] It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application.

[0043] ReferenceFigure 1 , Figure 1 The structural diagram of the photovoltaic module detection device related to the hardware running environment of the embodiment of the present application is shown.

[0044] As shown in Figure 1 , the photovoltaic module detection device can include a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to realize the connection and communication between the components. The user interface 1003 can include a display, an input unit such as a keyboard, and can also include a standard wired interface, a wireless interface. The network interface 1004 can optionally include a standard wired interface, a wireless interface (such as a wireless fidelity (Wi-Fi) interface). The memory 1005 can be a high-speed random access memory (RAM) memory, or a stable non-volatile memory (NVM), such as a disk memory. The memory 1005 can also be a storage device independent of the aforementioned processor 1001.

[0045] Those skilled in the art can understand that Figure 1 The structure shown in the figure does not constitute a limitation on the photovoltaic module detection device, and can include more or fewer components than the figure, or combine certain components, or different component arrangements.

[0046] As shown in Figure 1 , the memory 1005 as a storage medium can include an operating system, a network communication module, a user interface module, and a photovoltaic module detection program.

[0047] In the photovoltaic module detection device shown in Figure 1 , the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the photovoltaic module detection device of the present application can be arranged in the photovoltaic module detection device, and the photovoltaic module detection device calls the photovoltaic module detection program stored in the memory 1005 through the processor 1001, and executes the photovoltaic module detection method provided by the embodiment of the present application.

[0048] The embodiment of the present application provides a photovoltaic module detection method, which is described with reference to Figure 2 , Figure 2 The flowchart of the first embodiment of the photovoltaic module detection method of the present application is shown.

[0049] In this embodiment, the photovoltaic module detection method comprises the following steps:

[0050] Step S10: Obtain infrared image information according to the infrared camera.

[0051] It should be noted that the execution subject of the present embodiment is an unmanned flight vehicle, and the infrared camera is arranged on the unmanned flight vehicle. The unmanned flight vehicle can be a drone or other devices with the same or similar functions as the drone, and the present embodiment does not limit the same and takes the drone as an example to illustrate the present embodiment.

[0052] It should be noted that the present embodiment is applied to the inspection process of photovoltaic power stations and photovoltaic devices. At present, the operation and maintenance of specific photovoltaic modules mainly rely on manual inspection. However, large photovoltaic power stations are distributed in complex environments and have a huge coverage area. They are affected by terrain and present a disorderly and scattered nature. Therefore, it is very time-consuming and laborious to use manual inspection. For example, mountain power stations and water power stations bring great challenges to manual inspection. Therefore, the present embodiment proposes to use an unmanned flight vehicle for inspection because the unmanned flight vehicle has small terrain restrictions, a wide field of view, high efficiency, flexibility, and safety, which greatly facilitates the operation and maintenance of photovoltaic power stations. Further, the present embodiment proposes to use the infrared camera carried by the unmanned flight vehicle for imaging to quickly determine whether a fault occurs and the location and condition of the fault according to the temperature distribution of the photovoltaic module. This is because the infrared camera can accurately detect the surface temperature of the photovoltaic module without interfering with the operation of the power station. According to the surface temperature of the photovoltaic module, the fault can be detected, the safety hazard can be eliminated, and the impact of the fault on the power generation efficiency can be roughly estimated. Compared with a visible light camera, the infrared camera can detect faults caused by internal defects of the photovoltaic module. However, the visible light camera can only detect visible faults such as obstructions and glass breakage. Therefore, using the unmanned flight vehicle carrying the infrared camera to inspect the photovoltaic power station can improve the inspection efficiency, improve the safety, reduce the operation and maintenance cost, and greatly reduce the pressure on the operation and maintenance of the power station, which is of great significance to the stable operation of the photovoltaic power station.

[0053] It should be noted that the photovoltaic module mainly refers to a solar panel, a power transformer, an energy storage device, and an energy transmission device, and the present embodiment does not limit the same.

[0054] It can be understood that the infrared image information is the image obtained by rendering the temperature information collected by the infrared camera. Generally, the sensor of the infrared camera directly detects temperature information, and then the temperature information is rendered into an image through an algorithm. There are many rendering modes. If the lava mode is used, the rendering algorithm will use a similar equalization method to render the currently collected image information.

[0055] It should be noted that when the photovoltaic assembly, for example, solar panel, is faulty, the fault area is obviously different from the surrounding normal working area, for example, the solar panel surface is damaged or covered by an obstacle, and then the heat absorption efficiency at the position changes obviously, and a hot spot appears in the infrared image, and for another example, a defect or a circuit fault occurs in the photovoltaic assembly, and a hot spot also appears in the infrared image.

[0056] In the embodiment, the image quality of the infrared image is detected, and a flight adjustment instruction value is generated when the image quality of the infrared image is lower than a preset image quality; and the flight parameter of the unmanned aerial vehicle is adjusted according to the flight adjustment instruction until the image quality is no longer lower than the preset image quality.

[0057] It should be noted that the image quality indicates whether the collected infrared image is clear, and due to the influence of the environment, the distance and the device model, the collected image clarity may not reach the standard for identification, because the infrared camera is not accurately focused, and the flight speed of the unmanned aerial vehicle is too fast, and the like, which can cause the infrared image to be blurred and interfere with the photovoltaic assembly and the hot spot detection. The method of adding image quality judgment after the unmanned aerial vehicle collects the image is adopted to solve the problem.

[0058] In the embodiment, the flight height of the unmanned aerial vehicle is adjusted according to the flight adjustment instruction; and when the flight height adjustment of the unmanned aerial vehicle is completed and the image quality is lower than the preset image quality, the flight speed of the unmanned aerial vehicle is adjusted until the image quality is no longer lower than the preset image quality.

[0059] In the specific implementation, the image quality judgment is performed online after the unmanned aerial vehicle collects the image, if the image is blurred, the flight height of the unmanned aerial vehicle is first automatically adjusted, and then the flight speed of the unmanned aerial vehicle is automatically adjusted, until the image collected by the unmanned aerial vehicle is clear, and then the next flight and image collection and processing are continued.

[0060] Further, the embodiment provides a preferred scheme of flight control, to ensure that the unmanned aerial vehicle can reach the required image quality as soon as possible, for example, the flight adjustment instruction adjusts the flight height of the unmanned aerial vehicle, the unmanned aerial vehicle first flies to a preset first height from the ground, and gradually reduces to a second height from the ground, and during the descending process, the image quality meets the requirement and the process is stopped. If the image is still blurred when reaching the second height from the ground, the flight speed is gradually reduced until the image is clear.

[0061] Step S20: determining a target photovoltaic assembly image according to the infrared image information.

[0062] It should be noted that the photovoltaic module needs to be identified before the hot spot detection in the inspection process, so the photovoltaic module in the environmental image can be detected by using the deep learning neural network to lock the image in the target frame of the photovoltaic module, and the image in the target frame is the target photovoltaic module image.

[0063] In a specific implementation, the deep learning neural network for detecting photovoltaic modules can be trained by image samples with photovoltaic modules and environmental images without photovoltaic modules. During the training, a photovoltaic module is labeled as a positive sample as a whole, and the useless environmental image is trained as a negative sample.

[0064] Step S30: performing target detection on the target photovoltaic module image to determine the target hot spot.

[0065] It should be noted that after the target photovoltaic module image is determined, the fault point needs to be further confirmed. In different infrared images, the same temperature may correspond to different colors. It can be found by observing the unmanned aerial vehicle inspection video that when a very bright area appears in the field of view of the infrared camera, other areas will be darkened to a certain extent.

[0066] In this embodiment, the target detection model is used for identification. For example, a deep learning neural network composed of 12 residual components and 1001 convolution kernels is used for photovoltaic module and hot spot detection. The basic framework of the deep learning neural network can be divided into four parts: Input, Backbone, Neck, and Prediction. The Input part enriches the data set by splicing data enhancement, has low requirements for hardware devices, and low calculation cost. The Backbone part is mainly composed of CSP modules, and the features are extracted by CSPDarknet53. In the Neck, FPN and path aggregation network (PANet) are used to aggregate the image features at this stage. Finally, the network performs target prediction and outputs the prediction. The above neural network model is only used to illustrate the structure of the neural network, and does not limit the present scheme.

[0067] It can be understood that the deep learning neural network can detect a photovoltaic module and detect the position of the hot spot in the photovoltaic module. The two functions can be combined and performed in two steps: first, detecting the photovoltaic module, and then detecting the hot spot on the photovoltaic module. In this way, the photovoltaic module with the hot spot and the position of the hot spot can be detected.

[0068] In this embodiment, the unmanned aerial vehicle positioning information is obtained; the relative position information of the target hot spot is monitored; and the geographic position of the target photovoltaic module is determined according to the unmanned aerial vehicle positioning information and the relative position information.

[0069] It should be noted that after detecting the hot spot, the corresponding unmanned aerial vehicle GPS positioning information of the image can be used to locate which photovoltaic module has the hot spot, and then further corresponding processing is performed. Because after the target hot spot is confirmed, information related to the target hot spot needs to be obtained to provide information basis for subsequent maintenance. First, the specific location information of the hot spot needs to be determined, at which time the GPS positioning system of the unmanned aerial vehicle can be used to determine the relatively accurate geographic location. Through the GPS, the approximate location of the photovoltaic module with problems can be determined, and further it can be known which photovoltaic station has problems so that the maintenance personnel can go there. But even if the approximate location is known, the specific location of the photovoltaic module still needs to be identified according to the actual image captured by the unmanned aerial vehicle in general. In view of this situation, the orientation information of the hot spot can be calculated from the image collected by the unmanned aerial vehicle, and the relatively accurate hot spot location can be determined according to the current position of the unmanned aerial vehicle and the relative position (distance and azimuth) of the hot spot and the unmanned aerial vehicle.

[0070] In the embodiment, the pixel area of the target hot spot is determined, the pixel area of the photovoltaic module corresponding to the target hot spot is determined, the actual area of the target hot spot is determined according to the pixel area of the target hot spot, the pixel area of the photovoltaic module corresponding to the target hot spot, and the preset photovoltaic module product information, and the failure condition of the corresponding photovoltaic module is determined according to the actual area of the target hot spot.

[0071] It should be noted that after the target hot spot is confirmed, the situation of the hot spot needs to be further confirmed to help the operation and maintenance personnel prepare the tools and strategies for maintenance. Specifically, after detecting the photovoltaic module and the hot spot on the photovoltaic module, the ratio of the target frame area of the hot spot to the target frame area of the photovoltaic module can be calculated, which can be used as a reference value for the size of the hot spot area, and then help the maintenance personnel to speculate the fault type and scale. Since this step also needs to confirm the frame or surface of various components as a reference, it is necessary to collect the size parameters of various components in advance, that is, the preset photovoltaic module product information. In addition, the photovoltaic module needs to be identified according to the target photovoltaic module image to determine the type of the reference, for example: a hot spot appears in a solar panel, the solar panel where the hot spot is located can be identified to determine its model or type, and then the frame size of the corresponding solar panel is found by combining the preset photovoltaic module product information, and then the size of the target hot spot is converted. Therefore, identifying the photovoltaic module also needs to label the photovoltaic module, and various types and sizes of samples need to be collected and labeled to train the neural network, so that various types of targets can be detected during detection.

[0072] The embodiment acquires infrared image information according to the infrared camera; determines a target photovoltaic module image according to the infrared image information; and performs target detection on the target photovoltaic module image to determine a target hot spot. In this way, the photovoltaic module hot spot is automatically identified by using the unmanned aerial vehicle for inspection and combining the infrared imaging technology and the target detection technology, and the operation and maintenance efficiency of the photovoltaic equipment is improved.

[0073] Reference Figure 3 , Figure 3 It is a flowchart of a second embodiment of a photovoltaic module detection method.

[0074] Based on the above first embodiment, the photovoltaic module detection method in the step S30 comprises:

[0075] Step S31: inputting the target photovoltaic module image into a preset target detection model to obtain a target detection result.

[0076] It should be noted that the detection of the hot spot needs to be recognized by using the preset target detection model which is pre-trained, wherein the preset target detection model can be a neural network model, and specifically can be a deep neural network for photovoltaic module and hot spot detection using 12 residual components and 1001 convolution kernels to form a deep learning neural network. The basic framework of the deep learning neural network can be divided into four parts: Input, Backbone, Neck and Prediction. The Input part is enriched by data augmentation through splicing to enrich the data set, and the hardware device requirement is low and the calculation cost is low. The Backbone part is mainly composed of a CSP module, and the feature extraction is performed by CSPDarknet53. In the Neck, FPN and path aggregation network (PANet) are used to aggregate the image features at this stage. Finally, the network performs target prediction and obtains the final detection result through the prediction output, wherein the detection result includes whether there is a hot spot, and if there is, the position of the hot spot in the target photovoltaic module image is output.

[0077] Wherein, the deep learning neural network can detect a piece of photovoltaic module and also can detect which position of the photovoltaic module has a hot spot. The two functions can be combined and performed in two steps, i.e., detecting the photovoltaic module first and then detecting the hot spot on the photovoltaic module, so that the photovoltaic module with the hot spot and the position of the hot spot can be detected.

[0078] In the embodiment, training samples are acquired, and the training samples at least include normal photovoltaic module image samples, hot spot photovoltaic module image samples and reflective photovoltaic module image samples; a preset initial neural network model is trained according to the training samples to obtain a preset target detection model.

[0079] It should be noted that in the hot spot detection process, the most likely place for the neural network to misjudge is that the photovoltaic module may appear a reflection condition, when the photovoltaic module appears a reflected image of the sun, the other photovoltaic module area will be dark. In view of this phenomenon, when the hot spot and the reflection detection are carried out, the brightness contrast of the hot spot or the reflection area and the surrounding area and the shape information of the hot spot or the reflection area can be used. In the infrared image, the color development state of the reflection image and the color development state of the hot spot are very similar, the only difference is that the edge brightness change condition of the reflection image is more unobvious than that of the hot spot, that is, the hot spot has a more obvious boundary feeling, so in the model training process, the mixed sample of the photovoltaic module image sample with the reflection and the photovoltaic module image sample with the hot spot can be used as a difficult sample to train the neural network model to improve the distinguishing ability of the preset neural network model for the reflection image and the hot spot image. Because the solar panel actually exists the reflection phenomenon, the infrared image is similar to the hot spot, which brings interference. When the sample is labeled, we can distinguish the reflection and the hot spot according to the characteristics that the edge of the reflection spot is gradually changed, and then a kind of reflection is added in the sample training process. When the unmanned aerial vehicle detects the hot spot, the reflection and the hot spot are distinguished.

[0080] Step S32: determining a target hot spot according to the target detection result.

[0081] It can be understood that when the target result exists the hot spot information and the position of the hot spot detected, it can be known that the target photovoltaic exists the target hot spot, and after the hot spot is locked, other hot spot information can be further obtained according to the target hot spot.

[0082] In the embodiment, the target photovoltaic module image is input into a preset target detection model to obtain a target detection result, and a target hot spot is determined according to the target detection result. By adding difficult samples in the model training, the distinguishing degree of the hot spot and the reflection image is improved, the effective recognition of the hot spot is realized, and the accuracy of the hot spot recognition is improved.

[0083] In addition, the embodiment of the present application also provides a storage medium, and the storage medium stores a photovoltaic module detection program. When the photovoltaic module detection program is executed by a processor, the steps of the photovoltaic module detection method described above are realized.

[0084] Reference Figure 4 , Figure 4 The figure is a structural block diagram of the first embodiment of the photovoltaic module detection device of the present application.

[0085] As Figure 4 shown, the photovoltaic module detection device provided by the embodiment of the present application comprises:

[0086] The acquisition module 10 is configured to acquire infrared image information according to the infrared camera.

[0087] The processing module 20 is configured to determine a target photovoltaic component image according to the infrared image information.

[0088] The processing module 20 is further configured to determine a target hot spot by performing target detection on the target photovoltaic component image.

[0089] It should be understood that the above is only an example, and does not constitute any limitation on the technical solutions of the present application. In specific applications, those skilled in the art can set up as needed, and the present application does not limit this.

[0090] The processing module 20 is configured to determine a target photovoltaic component image according to the infrared image information. The processing module 20 is further configured to determine a target hot spot by performing target detection on the target photovoltaic component image. Through the above manner, the unmanned aerial vehicle is used for inspection, and the infrared imaging technology and the target detection technology are combined to automatically identify the hot spot of the photovoltaic component, thereby improving the operation and maintenance efficiency of the photovoltaic equipment.

[0091] In this embodiment, the processing module 20 is further configured to input the target photovoltaic component image into a preset target detection model to obtain a target detection result.

[0092] The target hot spot is determined according to the target detection result.

[0093] In this embodiment, the processing module 20 is further configured to obtain training samples, wherein the training samples at least include normal photovoltaic component image samples, hot spot photovoltaic component image samples, and reflective photovoltaic component image samples.

[0094] The preset initial neural network model is trained according to the training samples to obtain a preset target detection model.

[0095] In this embodiment, the processing module 20 is further configured to obtain unmanned aerial vehicle positioning information.

[0096] The relative position information of the target hot spot is monitored.

[0097] The geographic position of the target photovoltaic component is determined according to the unmanned aerial vehicle positioning information and the relative position information.

[0098] In this embodiment, the processing module 20 is further configured to determine the pixel area of the target hot spot.

[0099] The pixel area of the photovoltaic component corresponding to the target hot spot is determined.

[0100] According to the pixel area of the target hot spot, the pixel area of the corresponding photovoltaic module corresponding to the target hot spot, and preset photovoltaic module product information, a target hot spot actual area is determined, and a fault condition of the corresponding photovoltaic module is determined according to the target hot spot actual area.

[0101] In the embodiment, the processing module 20 is further configured to detect the image quality of the infrared image.

[0102] When the image quality of the infrared image is lower than a preset image quality, a flight adjustment instruction value is generated.

[0103] According to the flight adjustment instruction, the flight parameter of the unmanned aerial vehicle is adjusted until the image quality is no longer lower than the preset image quality.

[0104] In the embodiment, the processing module 20 is further configured to adjust the flight height of the unmanned aerial vehicle according to the flight adjustment instruction.

[0105] When the flight height adjustment of the unmanned aerial vehicle is completed and the image quality is lower than the preset image quality, the flight speed of the unmanned aerial vehicle is adjusted until the image quality is no longer lower than the preset image quality.

[0106] It should be noted that the above-described workflow is merely illustrative and does not limit the scope of protection of the present application. In actual application, a person skilled in the art can select part or all of the above-described workflow to achieve the purpose of the embodiment, and the present application is not limited in this regard.

[0107] In addition, technical details not described in detail in the embodiment can be found in the photovoltaic module detection method provided by any embodiment of the present application, and will not be described here.

[0108] In addition, it should be noted that in this document, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or system including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or system. Without more limitations, the element defined by the statement "includes a" does not exclude the presence of another identical element in the process, method, article or system including the element.

[0109] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0110] Those skilled in the art can clearly understand the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, can also be through hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application essentially or say the part of the prior art contribution can be embodied in the form of software products, the computer software product is stored in a storage medium (such as read only memory (Read Only Memory, ROM) / RAM, disk, optical disk), including a number of instructions to make a terminal device (may be a mobile phone, computer, server, or network equipment, etc.) executes the method described in various embodiments of the present application.

[0111] The above is only the preferred embodiment of the present application, not therefore limit the patent scope of the present application, any equivalent structure or equivalent flow transformation made by using the content of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for testing photovoltaic modules, characterized in that, The photovoltaic modules are inspected using a drone equipped with an infrared camera. The photovoltaic module inspection method includes: Infrared image information is acquired from the infrared camera; The target photovoltaic module image is determined based on the infrared image information; Target detection is performed on the image of the target photovoltaic module to identify the target hot spot; The step of performing target detection on the target photovoltaic module to determine the target hot spot includes: The target photovoltaic module image is input into a preset target detection model to obtain the target detection result; The target hot spot is determined based on the target detection results; Before inputting the target photovoltaic module image into a preset target detection model to obtain the target detection result, the method further includes: Acquire training samples, which include at least normal photovoltaic module image samples, hot spot photovoltaic module image samples, and reflective photovoltaic module image samples; The preset initial neural network model is trained based on the training samples to obtain the preset target detection model. During the model training process, the mixed samples with reflective photovoltaic module image samples and hot spot photovoltaic module image samples are used as difficult samples to train the neural network model in order to improve the preset neural network model's ability to distinguish between reflective images and hot spot images. After performing target detection on the target photovoltaic module image and determining the target hot spot, the method further includes: Determine the pixel area of ​​the target hot spot; Determine the pixel area of ​​the photovoltaic module corresponding to the target hot spot; The actual area of ​​the target hot spot is determined based on the pixel area of ​​the target hot spot, the pixel area of ​​the photovoltaic module corresponding to the target hot spot, and the preset photovoltaic module product information. The fault status of the corresponding photovoltaic module is determined based on the actual area of ​​the target hot spot. The photovoltaic module is identified based on the target photovoltaic module image to determine its type or model; the size parameters of the target photovoltaic module are determined by matching the type or model with preset photovoltaic module product information. The size parameters are used to calculate the actual area of ​​the target hot spot.

2. The method as described in claim 1, characterized in that, After performing target detection on the target photovoltaic module image and determining the target hot spot, the method further includes: Obtain drone location information; Monitor the relative position information of the target hot spot; The geographical location of the target photovoltaic module is determined based on the drone's positioning information and the relative position information.

3. The method as described in claim 1, characterized in that, After acquiring infrared image information from the infrared camera, the process further includes: Detecting the image quality of infrared images; When the image quality of the infrared image is lower than the preset image quality, a flight adjustment command value is generated; The flight parameters of the UAV are adjusted according to the flight adjustment command until the image quality is no longer lower than the preset image quality.

4. The method as described in claim 3, characterized in that, The step of adjusting the flight parameters of the UAV according to the flight adjustment command until the image quality is no longer lower than the preset image quality includes: Adjust the flight altitude of the UAV according to the flight adjustment command; When the drone's flight altitude is adjusted and the image quality is lower than the preset image quality, the drone's flight speed is adjusted until the image quality is no longer lower than the preset image quality.

5. A photovoltaic module testing device, characterized in that, The photovoltaic module testing device includes: The acquisition module is used to acquire infrared image information from the infrared camera; The processing module is used to determine the target photovoltaic module image based on the infrared image information; The processing module is also used to perform target detection on the target photovoltaic module image to determine the target hot spot; The processing module is further configured to input the image of the target photovoltaic module into a preset target detection model to obtain the target detection result; and determine the target hot spot based on the target detection result. The processing module is further configured to acquire training samples, which include at least normal photovoltaic module image samples, hot spot photovoltaic module image samples, and reflective photovoltaic module image samples; and to train a preset initial neural network model based on the training samples to obtain a preset target detection model. During the model training process, mixed samples containing reflective photovoltaic module image samples and hot spot photovoltaic module image samples are used as difficult samples to train the neural network model in order to improve the preset neural network model's ability to distinguish between reflective images and hot spot images. The processing module is further configured to: determine the pixel area of ​​the target hot spot; determine the pixel area of ​​the photovoltaic module corresponding to the target hot spot; determine the actual area of ​​the target hot spot based on the pixel area of ​​the target hot spot, the pixel area of ​​the photovoltaic module corresponding to the target hot spot, and preset photovoltaic module product information; determine the fault status of the corresponding photovoltaic module based on the actual area of ​​the target hot spot; identify the photovoltaic module based on the image of the target photovoltaic module to determine its type or model; and determine the size parameters of the target photovoltaic module by matching the type or model with the preset photovoltaic module product information; the size parameters are used to calculate the actual area of ​​the target hot spot.

6. A photovoltaic module testing device, characterized in that, The device includes: a memory, a processor, and a photovoltaic module testing program stored in the memory and executable on the processor, the photovoltaic module testing program being configured to implement the steps of the photovoltaic module testing method as described in any one of claims 1 to 4.

7. A storage medium, characterized in that, The storage medium stores a photovoltaic module testing program, which, when executed by a processor, implements the steps of the photovoltaic module testing method as described in any one of claims 1 to 4.

Images