Solar cell panel defect detection device and method based on non-contact ultrasonic imaging

Abstract

According to the solar cell panel defect detection device and method based on non-contact ultrasonic imaging, the non-contact ultrasonic imaging technology is adopted in the device, a clear target defect image of a solar cell panel to be detected can be obtained in a high-irradiance sunlight environment, and the defect detection accuracy is improved. Furthermore, the target defect image obtained by ultrasonic imaging can provide depth information, so that a surface defect detection result and an internal defect detection result of the solar cell panel to be detected can be identified, and comprehensive defect detection can be realized; invisible defects such as finer cracks and stress distribution can be detected through accurate ultrasonic signal acquisition, high-resolution image reconstruction and defect identification and marking processes, the high-precision detection requirement can be met, in addition, the device can still stably operate in complex environments such as strong light, high temperature or multiple noises, has good environmental adaptability, and can be widely applied to the field of ultrasonic detection. And full-coverage scanning of the solar cell panel to be detected can be realized, and the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of solar panels, and in particular, to a device and method for detecting defects in solar panels based on non-contact ultrasonic imaging. Background Art

[0002] There are often various invisible defects on the surface and inside of silicon crystal solar panels that are difficult to identify with the naked eye, including: hidden cracks, fragments, broken grids, stress, etc. Currently, the detection of invisible defects in solar panels mainly relies on electroluminescence imaging and photoluminescence imaging technologies. These two methods detect the electroluminescence or photoluminescence phenomena of solar panels, take images of the panels in low-light or dark environments, and identify their invisible defects, such as hidden cracks, fragments, broken grids, and stress concentrations. Specifically, electroluminescence imaging makes the panel emit light by applying a certain current to the panel, while photoluminescence imaging uses an external light source to excite the panel to emit light. These methods can generate images with a certain resolution, revealing a lot of internal defect information that cannot be directly observed by the naked eye.

[0003] However, in practical applications, since solar panels are usually installed in outdoor environments, the detection needs to be carried out under strong daylight conditions during the day. However, the spectral energy of electroluminescence and photoluminescence technologies is concentrated in the short-wave infrared region, which highly overlaps with the daylight spectrum energy and is easily interfered by ambient daylight. This interference will significantly reduce the image quality, making it difficult to accurately extract defect information. In addition, there are also certain limitations in the sensitivity of existing technologies to tiny defects. Especially in a high-irradiance environment, the weak characteristics of defect signals may be masked by interference signals, thereby reducing the reliability of detection.

[0004] The solutions for detecting defects in solar panels implemented by the above traditional electroluminescence imaging and photoluminescence imaging technologies have the following objective disadvantages:

[0005] Serious influence of ambient light interference: The traditional electroluminescence imaging and photoluminescence imaging technologies are very sensitive to ambient light conditions. Under high-irradiance daylight conditions, these methods cannot effectively shield external interference light sources, resulting in low-quality defect images being captured. The strong interference of ambient light not only masks the subtle features of defects but also introduces a large amount of noise, making it difficult for the detection system to accurately locate and identify invisible defects in solar panels;

[0006] Lack of depth information: Traditional electroluminescence and photoluminescence imaging mainly focuses on imaging the surface features of defects, lacking the ability to detect the depth information of defects, and it is difficult to identify hidden defects such as tiny cracks or stress distributions inside solar panels. This limits the comprehensiveness and accuracy of detection;

[0007] Insufficient defect detection sensitivity: Due to the weak spectral energy of defect feature information in the short infrared region, traditional imaging methods have problems with insufficient sensitivity in detecting tiny cracks, stress concentrations, and debris. This technical limitation is particularly evident when detecting more complex and subtle defects, and it is difficult to meet the high-precision detection requirements of practical applications;

[0008] Limitations in application scenarios: Traditional methods usually require operation in a dark environment, which poses strict requirements on the detection environment and time and cannot adapt to outdoor all-weather use scenarios. For example, the panels of large-scale power plants usually need to be quickly detected during the day, and traditional technologies cannot meet this requirement, significantly limiting the convenience and efficiency of practical use;

[0009] Low detection efficiency: Electroluminescence imaging and photoluminescence imaging require specific excitation conditions, which makes their detection processes relatively complex and difficult to complete the efficient detection of large-area panels in a short time. At the same time, the resolution of the generated images is limited by the device performance, which may lead to missed or misjudged defect information.

[0010] In summary, the defect detection of solar panels achieved by traditional electroluminescence imaging and photoluminescence imaging technologies has the following technical problems: being severely affected by ambient light, insufficient defect information, insufficient sensitivity of defect detection, limited application scenarios, and low detection efficiency. Summary of the Invention

[0011] In view of this, the purpose of the present invention is to provide a defect detection device and method for solar panels based on non-contact ultrasonic imaging to alleviate the technical problems of traditional solar panel defect detection methods, such as being severely affected by ambient light, insufficient defect information, insufficient sensitivity of defect detection, limited application scenarios, and low detection efficiency.

[0012] In the first aspect, an embodiment of the present invention provides a defect detection device for solar panels based on non-contact ultrasonic imaging, including: an ultrasonic detection module, an image reconstruction module, and a control and analysis module;

[0013] The ultrasonic detection module is used to emit ultrasonic signals with a preset frequency to the solar panel to be detected placed below it, and perform full-coverage scanning on the solar panel to be detected, so as to obtain the scanned ultrasonic signals generated by the interaction between the solar panel to be detected and the ultrasonic signals, and then convert the scanned ultrasonic signals into electrical signals. Among them, the scanned ultrasonic signals carry defect information of the solar panel to be detected, and the scanned ultrasonic signals include: reflected ultrasonic signals and / or scattered ultrasonic signals and / or transmitted ultrasonic signals;

[0014] The image reconstruction module is used to reconstruct the electrical signals sent by the ultrasonic detection module into an initial defect image, and perform high-resolution reconstruction on the initial defect image by using a high-resolution image reconstruction network to obtain a target defect image;

[0015] The control and analysis module is used to perform defect recognition and annotation on the target defect image by using a defect recognition network to obtain the defect detection result of the solar panel to be detected. Among them, the defect detection result includes: the type of defects on the surface and inside of the solar panel to be detected, the position of the defects on the surface and inside of the solar panel to be detected, and the size of the defects on the surface and inside of the solar panel to be detected.

[0016] Furthermore, the ultrasonic detection module includes: an ultrasonic transmitter, an ultrasonic scanner, and an ultrasonic sensor;

[0017] The ultrasonic transmitter is used to emit ultrasonic signals with a preset frequency to the solar panel to be detected placed below it. Among them, the distance between the ultrasonic transmitter and the solar panel to be detected can make the ultrasonic signals emitted by the ultrasonic transmitter evenly cover the surface and inside of the solar panel to be detected;

[0018] The ultrasonic scanner is used to perform full-coverage scanning on the solar panel to be detected, and scan to obtain the scanned ultrasonic signals generated by the interaction between the ultrasonic signals and the solar panel to be detected;

[0019] The ultrasonic sensor is used to receive the scanned ultrasonic signals sent by the ultrasonic scanner and convert the scanned ultrasonic signals into electrical signals.

[0020] Furthermore, the control and analysis module is also used to control the ultrasonic transmitter, the ultrasonic scanner, the ultrasonic sensor, and the image reconstruction module to work.

[0021] Further, the high-resolution image reconstruction network includes: a feature extraction module, a depth self-attention module, a spatial upsampling module, and a feature mapping module connected in sequence, wherein the input of the feature extraction module is also connected to the depth self-attention module in an additive manner, and the number of the depth self-attention modules is multiple.

[0022] Further, the depth self-attention module includes: a normalization layer, a depth self-attention layer, a normalization layer, and a depth feed-forward network, wherein the depth self-attention layer includes: a first depth convolution and normalization, and a second depth convolution and normalization.

[0023] Further, the defect recognition network includes: a feature extraction module, a first feature mapping module, a depth self-attention module, a second feature mapping module, and a defect recognition module connected in sequence.

[0024] Further, the ultrasonic wave emission source adopts an array design.

[0025] Further, the ultrasonic scanner includes: a mechanical scanner or an electronic scanner.

[0026] Further, the ultrasonic scanner includes: a multi-axis movable platform and a high-precision sensor.

[0027] In a second aspect, an embodiment of the present invention further provides a method for detecting defects in a solar panel based on non-contact ultrasonic imaging, which is applied to the device for detecting defects in a solar panel based on non-contact ultrasonic imaging according to any one of the above first aspects. The method includes:

[0028] Emitting an ultrasonic wave signal with a preset frequency to the solar panel to be detected, and performing full-coverage scanning on the solar panel to be detected to obtain a scanned ultrasonic wave signal generated by the interaction between the solar panel to be detected and the ultrasonic wave signal, and then converting the scanned ultrasonic wave signal into an electrical signal, wherein the scanned ultrasonic wave signal carries defect information of the solar panel to be detected, and the scanned ultrasonic wave signal includes: a reflected ultrasonic wave signal and / or a scattered ultrasonic wave signal and / or a transmitted ultrasonic wave signal;

[0029] Reconstructing the electrical signal into an initial defect image, and performing high-resolution reconstruction on the initial defect image by using a high-resolution image reconstruction network to obtain a target defect image;

[0030] Use a defect recognition network to perform defect recognition and annotation on the target defect image, and obtain the defect detection result of the solar panel to be detected, where the defect detection result includes: the types of defects on the surface and inside of the solar panel to be detected, the positions of the defects on the surface and inside of the solar panel to be detected, and the sizes of the defects on the surface and inside of the solar panel to be detected.

[0031] In an embodiment of the present invention, a solar panel defect detection device based on non-contact ultrasonic imaging is provided, including: an ultrasonic detection module, an image reconstruction module, and a control and analysis module; the ultrasonic detection module is configured to emit ultrasonic signals with a preset frequency to a solar panel to be detected placed below it, and perform full-coverage scanning on the solar panel to be detected, obtain scanning ultrasonic signals generated by the interaction between the solar panel to be detected and the ultrasonic signals, and then convert the scanning ultrasonic signals into electrical signals, wherein the scanning ultrasonic signals carry defect information of the solar panel to be detected, and the scanning ultrasonic signals include: reflected ultrasonic signals and / or scattered ultrasonic signals and / or transmitted ultrasonic signals; the image reconstruction module is configured to reconstruct the electrical signals sent by the ultrasonic detection module into an initial defect image, and perform high-resolution reconstruction on the initial defect image by using a high-resolution image reconstruction network to obtain a target defect image; the control and analysis module is configured to perform defect recognition and annotation on the target defect image by using a defect recognition network to obtain a defect detection result of the solar panel to be detected, wherein the defect detection result includes: the types of defects on the surface and inside of the solar panel to be detected, the positions of the defects on the surface and inside of the solar panel to be detected, and the sizes of the defects on the surface and inside of the solar panel to be detected. From the above description, it can be seen that in the solar panel defect detection device based on non-contact ultrasonic imaging of the present invention, the non-contact ultrasonic imaging technology is adopted to avoid the influence of sunlight spectrum energy on the imaging result, and a clear target defect image of the solar panel to be detected can be obtained in a high-irradiance sunlight environment. Moreover, the target defect image obtained by ultrasonic imaging can provide depth information, and thus the defect detection results on the surface and inside of the solar panel to be detected can be identified, that is, comprehensive defect detection can be realized. In addition, accurate ultrasonic signal acquisition, high-resolution image reconstruction, and defect recognition and annotation processes can detect more subtle invisible defects such as cracks and stress distributions, which can meet the requirements of high-precision detection, that is, the detection sensitivity is high. In addition, it can still operate stably in complex environments such as strong light, high temperature, or multi-noise, and has good environmental adaptability, that is, the application scenario is wide. It can also perform full-coverage scanning of the solar panel to be detected, improve the detection efficiency, and be applicable to industrial batch detection, alleviating the technical problems of traditional solar panel defect detection methods being severely affected by ambient light, the defect information being incomplete, the defect detection sensitivity being insufficient, the application scenario being limited, and the detection efficiency being low. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0033] Figure 1 Schematic diagram of a solar panel defect detection device based on non-contact ultrasonic imaging provided by an embodiment of the present invention;

[0034] Figure 2 Schematic diagram of the structure of a high-resolution image reconstruction network provided by an embodiment of the present invention;

[0035] Figure 3 Schematic diagram of the structure of a depth self-attention module provided by an embodiment of the present invention;

[0036] Figure 4 Schematic diagram of the structure of a depth self-attention layer provided by an embodiment of the present invention;

[0037] Figure 5 Schematic diagram of the structure of a defect recognition network provided by an embodiment of the present invention;

[0038] Figure 6 Schematic diagram of a method for detecting solar panel defects based on non-contact ultrasonic imaging provided by an embodiment of the present invention. Specific Embodiments

[0039] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0040] Traditional solar panel defect detection methods are severely affected by ambient light, the defect information is not comprehensive enough, the sensitivity of defect detection is insufficient, the application scenarios are limited, and the detection efficiency is low.

[0041] Based on this, in the solar panel defect detection device based on non-contact ultrasonic imaging of the present invention, the non-contact ultrasonic imaging technology is adopted, which avoids the influence of sunlight spectral energy on the imaging result, can obtain a clear target defect image of the solar panel to be detected in a high-irradiance sunlight environment, and the target defect image obtained by ultrasonic imaging can provide depth information, so that the defect detection results on the surface of the solar panel to be detected and the defect detection results inside the solar panel to be detected can be identified, that is, comprehensive defect detection can be realized. In addition, precise ultrasonic signal acquisition, high-resolution image reconstruction, and defect recognition and annotation processes can detect more subtle invisible defects such as cracks and stress distributions, which can meet the high-precision detection requirements, that is, the detection sensitivity is high. In addition, it can still operate stably in complex environments such as strong light, high temperature or multi-noise, and has good environmental adaptability, that is, it has a wide range of application scenarios. It can also achieve full coverage scanning of the solar panel to be detected, improve the detection efficiency, and is applicable to industrial batch detection.

[0042] For the convenience of understanding this embodiment, first, a solar panel defect detection device based on non-contact ultrasonic imaging disclosed in the embodiments of the present invention will be introduced in detail.

[0043] Embodiment 1:

[0044] Figure 1 It is a schematic diagram of a solar panel defect detection device based on non-contact ultrasonic imaging according to an embodiment of the present invention. As Figure 1 shown, the solar panel defect detection device based on non-contact ultrasonic imaging includes: an ultrasonic detection module, an image reconstruction module, and a control and analysis module;

[0045] The ultrasonic detection module is used to emit ultrasonic signals with a preset frequency to the solar panel to be detected placed below it, and perform full coverage scanning on the solar panel to be detected to obtain scanning ultrasonic signals generated by the interaction between the solar panel to be detected and the ultrasonic signals, and then convert the scanning ultrasonic signals into electrical signals. Among them, the scanning ultrasonic signals carry defect information of the solar panel to be detected, and the scanning ultrasonic signals include: reflected ultrasonic signals and / or scattered ultrasonic signals and / or transmitted ultrasonic signals;

[0046] The image reconstruction module is used to reconstruct the electrical signals sent by the ultrasonic detection module into an initial defect image, and perform high-resolution reconstruction on the initial defect image using a high-resolution image reconstruction network to obtain a target defect image;

[0047] A control and analysis module is used to perform defect recognition and annotation on a target defect image by using a defect recognition network, so as to obtain a defect detection result of the solar panel to be detected. The defect detection result includes: the types of defects on the surface and inside of the solar panel to be detected, the positions of the defects on the surface and inside of the solar panel to be detected, and the sizes of the defects on the surface and inside of the solar panel to be detected.

[0048] In the embodiment of the present invention, the solar panel defect detection device based on non-contact ultrasonic imaging is a non-contact non-destructive testing device. The defect recognition network accurately obtains the defect detection result by extracting features such as the shape, size, and position of the defect area in the target defect image. The types of the above-mentioned defects include: hidden cracks, fragments, broken grids, etc.

[0049] In an embodiment of the present invention, a solar panel defect detection device based on non-contact ultrasonic imaging is provided, including: an ultrasonic detection module, an image reconstruction module, and a control and analysis module; the ultrasonic detection module is configured to emit ultrasonic signals with a preset frequency to a solar panel to be detected placed below it, and perform full-coverage scanning on the solar panel to be detected, obtain scanning ultrasonic signals generated by the interaction between the solar panel to be detected and the ultrasonic signals, and then convert the scanning ultrasonic signals into electrical signals, wherein the scanning ultrasonic signals carry defect information of the solar panel to be detected, and the scanning ultrasonic signals include: reflected ultrasonic signals and / or scattered ultrasonic signals and / or transmitted ultrasonic signals; the image reconstruction module is configured to reconstruct the electrical signals sent by the ultrasonic detection module into an initial defect image, and perform high-resolution reconstruction on the initial defect image by using a high-resolution image reconstruction network to obtain a target defect image; the control and analysis module is configured to perform defect recognition and annotation on the target defect image by using a defect recognition network to obtain a defect detection result of the solar panel to be detected, wherein the defect detection result includes: the types of defects on the surface and inside of the solar panel to be detected, the positions of the defects on the surface and inside of the solar panel to be detected, and the sizes of the defects on the surface and inside of the solar panel to be detected. Through the above description, it can be seen that in the solar panel defect detection device based on non-contact ultrasonic imaging of the present invention, the non-contact ultrasonic imaging technology is adopted to avoid the influence of sunlight spectrum energy on the imaging result, and a clear target defect image of the solar panel to be detected can be obtained under a high-irradiance sunlight environment. Moreover, the target defect image obtained by ultrasonic imaging can provide depth information, and then the defect detection results on the surface and inside of the solar panel to be detected can be identified, that is, comprehensive defect detection can be realized. In addition, accurate ultrasonic signal acquisition, high-resolution image reconstruction, and defect recognition and annotation processes can detect more subtle cracks, stress distributions and other invisible defects, which can meet the high-precision detection requirements, that is, the detection sensitivity is high. In addition, it can still operate stably under complex environments such as strong light, high temperature or high noise, and has good environmental adaptability, that is, the application scenario is wide. It can also perform full-coverage scanning of the solar panel to be detected, improve the detection efficiency, be applicable to industrial batch detection, and alleviate the technical problems of the traditional solar panel defect detection method being seriously affected by environmental light, the defect information being incomplete, the defect detection sensitivity being insufficient, the application scenario being limited, and the detection efficiency being low.

[0050] The above content briefly introduces the solar panel defect detection device based on non-contact ultrasonic imaging of the present invention, and the following will describe the specific content involved in detail.

[0051] In an optional embodiment of the present invention, with reference to Figure 1,The ultrasonic detection module includes: an ultrasonic transmitting source, an ultrasonic scanner and an ultrasonic sensor;

[0052] An ultrasonic emission source, used to emit an ultrasonic signal of a preset frequency to a solar panel to be detected placed thereunder, wherein the distance between the ultrasonic emission source and the solar panel to be detected is such that the ultrasonic signal emitted by the ultrasonic emission source uniformly covers the surface and interior of the solar panel to be detected;

[0053] An ultrasonic scanner is used to perform a full coverage scan on the solar panel to be inspected, and obtain a scanning ultrasonic signal generated by the interaction between the ultrasonic signal and the solar panel to be inspected;

[0054] The ultrasonic sensor is used to receive the scanning ultrasonic signal sent by the ultrasonic scanner and convert the scanning ultrasonic signal into an electrical signal.

[0055] When implemented, the ultrasonic emission source emits an ultrasonic signal of a specific frequency (the ultrasonic signal of the specific frequency may be a high-frequency ultrasonic signal) to ensure that the ultrasonic signal covers the solar panel to be detected (which may be 2m 2 The surface and internal structure of the solar panel to be inspected can be measured by the non-contact design, which eliminates the need for direct contact with the solar panel to be inspected, thus avoiding potential damage to the solar panel to be inspected by traditional methods.

[0056] Ultrasonic scanners can detect large-area solar panels (e.g. 2m 2 ) is fully covered. During the scanning process, the ultrasonic signal interacts with the internal structure of the solar panel to be inspected, generating reflected ultrasonic waves, scattered ultrasonic waves or transmitted ultrasonic waves, collectively referred to as scanning ultrasonic signals. These scanning ultrasonic signals contain the physical characteristics and defect information of the solar panel to be inspected.

[0057] The ultrasonic sensor converts the received scanning ultrasonic signal sent by the ultrasonic scanner into an electrical signal and sends it to the image reconstruction module. The ultrasonic sensor supports a wide frequency response range and can detect tiny signal changes.

[0058] In an optional embodiment of the present invention, the control and analysis module is also used to control the operation of the ultrasonic emission source, the ultrasonic scanner, the ultrasonic sensor and the image reconstruction module.

[0059] Specifically, the control and analysis module integrates core algorithms for equipment control (controlling the ultrasonic emission source, ultrasonic scanner, ultrasonic sensor and image reconstruction module), signal processing and defect identification, and can process and analyze the collected data in real time.

[0060] In an alternative embodiment of the present invention, the high-resolution image reconstruction network includes: a feature extraction module, a depth self-attention module, a spatial upsampling module, and a feature mapping module connected in sequence. Among them, the input of the feature extraction module is also connected in an additive manner to the depth self-attention module, and the number of depth self-attention modules is multiple.

[0061] Specifically, the above high-resolution image reconstruction network can highlight the tiny cracks, stress concentration areas, and other invisible defects in the solar panel to be detected. That is, the obtained target defect image can reveal the invisible defects on the surface and inside of the solar panel to be detected.

[0062] The structure of the above network is as Figure 2 shown. Specifically, referring to Figure 3 and Figure 4 the depth self-attention module includes: a normalization layer, a depth self-attention layer, a normalization layer, and a depth feed-forward network. Among them, the depth self-attention layer includes: a first depth convolution and normalization, and a second depth convolution and normalization.

[0063] Specifically, after multiplying the outputs of the first depth convolution and normalization by themselves, the product result, the product result, and the output result of themselves are multiplied and then input into the second depth convolution and normalization.

[0064] In an alternative embodiment of the present invention, referring to Figure 5 the defect recognition network includes: a feature extraction module, a first feature mapping module, a depth self-attention module, a second feature mapping module, and a defect recognition module connected in sequence.

[0065] Specifically, the above feature extraction module, first feature mapping module, and depth self-attention module can further extract depth information, remove environmental noise, and improve the accuracy and reliability of defect detection; the above second feature mapping module is a convolutional decoding network for gradually restoring the feature map; the above defect recognition module is a classification prediction head composed of a two-dimensional convolutional layer and a Sigmoid activation function, which classifies the features to generate the final defect detection result.

[0066] In an alternative embodiment of the present invention, the ultrasonic wave emission source adopts an array design.

[0067] Specifically, the above array design can cover a solar panel of 2m 2 in size.

[0068] In an alternative embodiment of the present invention, the ultrasonic scanner includes: a mechanical scanner or an electronic scanner.

[0069] Specifically, the ultrasonic scanner includes a multi-axis movable platform and a high-precision sensor. The design of the multi-axis movable platform enables full-coverage scanning of the solar panel to be detected. Combined with the high-precision sensor, it can ensure the smoothness of movement and the accuracy of positioning during the imaging process.

[0070] The device of the present invention mainly includes the following technical features:

[0071] 1. Strong anti-environmental light interference ability: Adopting non-contact ultrasonic imaging technology, it can completely avoid the interference of environmental sunlight, without the need to operate under low light or dark room conditions, and realizes clear defect imaging in a high-irradiance sunlight environment, greatly improving the detection reliability;

[0072] 2. Ability to detect depth information: Compared with traditional electroluminescence and photoluminescence imaging methods that can only capture surface defect information, the present invention realizes the detection of depth information of internal defects (such as hidden cracks and stress distribution) in solar panels through ultrasonic imaging technology, and can comprehensively reflect the internal structure status of the panels;

[0073] 3. High-sensitivity defect detection: The ultrasonic imaging method of the present invention has high detection sensitivity, can effectively capture tiny defects such as small cracks, broken grids, and fragments, and significantly improves the detection accuracy of invisible defects;

[0074] 4. Large-area and rapid scanning imaging: The device is equipped with an efficient ultrasonic scanner, which can cover a large area (such as 2m 2 ) of solar panels, significantly improving the detection efficiency and meeting the requirements of rapid detection of large-scale components in photovoltaic power stations;

[0075] 5. Adaptive imaging reconstruction algorithm (i.e., high-resolution image reconstruction network): Through an innovative ultrasonic high-resolution image reconstruction algorithm, combined with multi-dimensional feature extraction and depth information processing, it can accurately reproduce the spatial distribution and internal characteristics of defects, providing high-quality target defect images for subsequent maintenance;

[0076] 6. Wide applicability: It can work in a high-irradiance sunlight environment and other complex detection conditions, without the need for a special detection environment, greatly expanding its application scenarios, including: outdoor photovoltaic power stations, production workshops and other occasions.

[0077] Embodiment 2:

[0078] Figure 6 It is a schematic diagram of a method for detecting defects in solar panels based on non-contact ultrasonic imaging according to an embodiment of the present invention. As Figure 6 shown, the method includes:

[0079] Step S602: Transmit an ultrasonic signal with a preset frequency to the solar panel to be detected, and perform full-coverage scanning on the solar panel to be detected to obtain a scanned ultrasonic signal generated by the interaction between the solar panel to be detected and the ultrasonic signal. Then, convert the scanned ultrasonic signal into an electrical signal. The scanned ultrasonic signal carries defect information of the solar panel to be detected, and the scanned ultrasonic signal includes: reflected ultrasonic signal and / or scattered ultrasonic signal and / or transmitted ultrasonic signal;

[0080] Step S604: Reconstruct the electrical signal into an initial defect image, and perform high-resolution reconstruction on the initial defect image using a high-resolution image reconstruction network to obtain a target defect image;

[0081] Step S606: Use a defect recognition network to perform defect recognition and annotation on the target defect image to obtain a defect detection result of the solar panel to be detected. The defect detection result includes: the type of defects on the surface and inside of the solar panel to be detected, the location of the defects on the surface and inside of the solar panel to be detected, and the size of the defects on the surface and inside of the solar panel to be detected.

[0082] The process of defect detection is briefly described below:

[0083] (1) Pre-scanning preparation

[0084] Place the solar panel to be detected under the device, start the ultrasonic scanner, and adjust the distance between the ultrasonic transmitter and the solar panel to be detected to ensure that the ultrasonic wave evenly covers the solar panel to be detected.

[0085] (2) Ultrasonic wave emission and signal acquisition

[0086] The ultrasonic transmitter emits high-frequency ultrasonic signals at a certain frequency, and the ultrasonic signals cover the surface and inside of the solar panel to be detected.

[0087] The ultrasonic scanner receives the ultrasonic signals (i.e., scanned ultrasonic signals) reflected / scattered / transmitted by the solar panel to be detected. The ultrasonic sensor records the amplitude, frequency, and phase information of the scanned ultrasonic signal in real time (i.e., obtains an electrical signal).

[0088] (3) Image reconstruction and depth information extraction

[0089] Use a traditional image reconstruction algorithm to process the received electrical signal to obtain an initial defect image, and use a high-resolution image reconstruction network to convert the initial defect image into a high-resolution two-dimensional and three-dimensional target defect image.

[0090] During the reconstruction process, the depth information of the solar panel to be detected is extracted for subsequent identification of internal invisible defects of the solar panel to be detected, such as hidden cracks and stress distribution.

[0091] (4) Defect identification and classification

[0092] The control and analysis module combines a deep learning algorithm (i.e., a defect identification network) to analyze and classify the target defect image, distinguish different types of defects such as cracks, broken grids, and debris, and can provide an automatic annotation function to highlight the detected defect areas and output the defect type, specific location, and size.

[0093] The solar panel defect detection device and method based on non-contact ultrasonic imaging of the present invention have the following advantages:

[0094] 1) Not affected by ambient light: Using non-contact ultrasonic imaging technology, the influence of solar spectrum energy on the imaging result is avoided, and clear solar panel defect images can be obtained in a high-irradiance sunlight environment;

[0095] 2) High detection sensitivity: Through precise ultrasonic signal acquisition and image reconstruction algorithms (i.e., high-resolution image reconstruction network), more subtle invisible defects such as cracks and stress distribution can be detected, suitable for high-precision detection requirements;

[0096] 3) Having the ability to analyze depth information: Combining three-dimensional reconstruction technology (i.e., the process of reconstructing the initial defect image and high-resolution reconstruction), not only surface defect information is provided, but also the depth and internal distribution of the defects can be analyzed, thus achieving more comprehensive defect detection;

[0097] 4) Supporting large-area scanning: The ultrasonic scanner is flexibly designed to support scanning of solar panels with a range of 2m 2 or larger, improving the detection efficiency and being suitable for industrial batch detection;

[0098] 5) Adapting to complex detection environments: It can still operate stably in complex environments such as strong light, high temperature, or high noise, and has good environmental adaptability;

[0099] 6) Improving the accuracy of defect classification: Combining deep learning algorithms and signal analysis techniques to achieve automatic identification and classification of defect types, reducing manual intervention and improving the reliability and consistency of detection results;

[0100] 7) Non-destructive testing: Based on the non-contact detection method of ultrasonic waves, physical damage to the solar panel is avoided, making it suitable for long-term use of high-value components.

[0101] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems and devices described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0102] In addition, in the description of the embodiments of the present invention, unless otherwise clearly defined and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be a direct connection or an indirect connection through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0103] If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0104] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A solar panel defect detection device based on non-contact ultrasonic imaging, characterized in that: include: Ultrasonic detection module, image reconstruction module, control and analysis module; The ultrasonic detection module is used to transmit an ultrasonic signal of a preset frequency to a solar panel to be detected placed thereunder, and perform a full coverage scan on the solar panel to be detected to obtain a scanning ultrasonic signal generated by the interaction between the solar panel to be detected and the ultrasonic signal, and then convert the scanning ultrasonic signal into an electrical signal, wherein the scanning ultrasonic signal carries defect information of the solar panel to be detected, and the scanning ultrasonic signal includes: a reflected ultrasonic signal and / or a scattered ultrasonic signal and / or a transmitted ultrasonic signal; The image reconstruction module is used to reconstruct the electrical signal sent by the ultrasonic detection module into an initial defect image, and use a high-resolution image reconstruction network to perform high-resolution reconstruction on the initial defect image to obtain a target defect image; The control and analysis module is used to use a defect recognition network to perform defect recognition and annotation on the target defect image to obtain a defect detection result of the solar panel to be detected, wherein the defect detection result includes: types of defects on the surface and inside of the solar panel to be detected, locations of defects on the surface and inside of the solar panel to be detected, and sizes of defects on the surface and inside of the solar panel to be detected.

2. The device according to claim 1, characterized in that The ultrasonic detection module includes: an ultrasonic emission source, an ultrasonic scanner and an ultrasonic sensor; The ultrasonic emission source is used to emit an ultrasonic signal of a preset frequency to the solar panel to be detected placed thereunder, wherein the distance between the ultrasonic emission source and the solar panel to be detected is such that the ultrasonic signal emitted by the ultrasonic emission source uniformly covers the surface and interior of the solar panel to be detected; The ultrasonic scanner is used to perform a full coverage scan on the solar panel to be inspected, and obtain a scanning ultrasonic signal generated by the interaction between the ultrasonic signal and the solar panel to be inspected; The ultrasonic sensor is used to receive the scanning ultrasonic signal sent by the ultrasonic scanner and convert the scanning ultrasonic signal into an electrical signal.

3. The device according to claim 2, characterized in that The control and analysis module is also used to control the operation of the ultrasonic emission source, the ultrasonic scanner, the ultrasonic sensor and the image reconstruction module.

4. The device according to claim 1, characterized in that The high-resolution image reconstruction network includes: a feature extraction module, a deep self-attention module, a spatial upsampling module and a feature mapping module connected in sequence, wherein the input of the feature extraction module is also summed and connected with the deep self-attention module, and the number of the deep self-attention modules is multiple.

5. The device according to claim 4, characterized in that The deep self-attention module includes: a normalization layer, a deep self-attention layer, a normalization layer and a deep feedforward network, wherein the deep self-attention layer includes: a first deep convolution and normalization, a second deep convolution and normalization.

6. The device according to claim 1, characterized in that The defect recognition network includes: a feature extraction module, a first feature mapping module, a deep self-attention module, a second feature mapping module and a defect recognition module which are connected in sequence.

7. The device according to claim 2, characterized in that The ultrasonic wave emission source adopts an array design.

8. The device according to claim 2, characterized in that The ultrasonic scanner includes: a mechanical scanner or an electronic scanner.

9. The device according to claim 2, characterized in that The ultrasonic scanner comprises: a multi-axis movable platform and a high-precision sensor.

10. A method for detecting defects in solar panels based on non-contact ultrasonic imaging, characterized in that: The solar panel defect detection device based on non-contact ultrasonic imaging applied to any one of claims 1 to 9, the method comprising: Transmitting an ultrasonic signal of a preset frequency to a solar panel to be inspected, and performing a full coverage scan on the solar panel to be inspected, obtaining a scanning ultrasonic signal generated by the interaction between the solar panel to be inspected and the ultrasonic signal, and then converting the scanning ultrasonic signal into an electrical signal, wherein the scanning ultrasonic signal carries defect information of the solar panel to be inspected, and the scanning ultrasonic signal includes: a reflected ultrasonic signal and / or a scattered ultrasonic signal and / or a transmitted ultrasonic signal; Reconstructing the electrical signal into an initial defect image, and using a high-resolution image reconstruction network to perform high-resolution reconstruction on the initial defect image to obtain a target defect image; A defect recognition network is used to perform defect recognition and annotation on the target defect image to obtain a defect detection result of the solar panel to be detected, wherein the defect detection result includes: types of defects on the surface and inside of the solar panel to be detected, locations of defects on the surface and inside of the solar panel to be detected, and sizes of defects on the surface and inside of the solar panel to be detected.

Images