Description:FIELD OF INVENTION

[001] The present invention relates to detect defects in a solar panel. Specifically, the present invention relates to a system of detecting defects in the solar panel using electroluminescence (EL) images and visual images of the solar panel.

BACKGROUND OF THE INVENTION

[002] Electrical energy is being produced by utilizing solar emission through photovoltaic cells in the solar panels. The photovoltaic cells are needed to be taken care of for better life and performance. The photovoltaic cells tend to degrade with time which leads to decrease in the performance, lifetime and reliability of the photovoltaic cells. The performance of solar panels are also degraded by incorrect arrangement of solar cells, short circuits, scratches etc. Defects in the solar panels happen during assembly, manufacturing and during operation in the field.

[003] Detecting defects in the solar panels along with other defects in traditional method involve manual inspection. Each solar panel is checked by analyzing electroluminescence that is emitted when the solar panels are applied with current. Analyzing each solar module by electroluminescence test is more time taking and it may incur human error.

[004] The patent application US2019089301A1 titled “System and method for solar cell defect detection” discloses a system for testing a solar panel includes an electrical power supply, an imaging device, and a computing device. The electrical power supply is configured to couple to a solar panel and supply an electrical current in the solar panel, thereby inducing electroluminescence. The imaging device is configured to measure the electroluminescence of the solar panel. The computing device is coupled to the imaging device, the computing device configured to determine a defect in the solar panel based on a measurement of the electroluminescence.

[005] The above described patent has disadvantage that it does not provide a system with automating defect detection process, using advanced machine learning modules and framework to identify a wide range of defects with high precision. Enhancing the throughput of the inspection process, allowing for quicker and more reliable quality control, thereby reducing human labor and error.

[006] In order to overcome the problem associated with the state of arts, there is a need for the development of a system that overcomes the aforesaid limitations in a more efficient manner.

OBJECTIVE OF THE INVENTION
[007] The primary objective of the present invention is to provide a system to detect defects in a solar panel.

[008] Another objective of the present invention is to develop an automated system to detect multiple types of defects in photovoltaic cells of the solar panel such as dry soldering, latent defects and micro -cracks.

[009] Yet another objective of the present invention is to detect a wide range of defects such as broken cells, black borders, scratches and no electricity in solar cells of the solar panel through integration of electroluminescence (EL) and visual images.

[0010] Another objective of the present invention is to implement a real-time monitoring solution.

[0011] Yet another objective of the present invention is to integrate artificial intelligence (AI)-driven suggestion to initiate repairs.

[0012] Another objective of the present invention is to provide the system capable of detecting defects with an improved accuracy.

[0013] Yet another objective of the present invention is to provide an artificial intelligence (AI) powered repair suggestions, reducing time and increasing production efficiency.

[0014] Another objective of the present invention is to provide a method that does not damage the solar panel during testing.

[0015] Yet another objective of the present invention is to identify potential issues that cause failure in the solar panel during operation.

[0016] Another objective of the present invention is to detect defects at early stage in production, allowing timely rework.

[0017] Yet another objective of the present invention is to improve the robustness and reliability of defect detection by using diverse image data, ensuring a comprehensive analysis of defects in different stages of solar panel production.

[0018] Other objectives and advantages of the present invention will become apparent from the following description taken in connection with the accompanying drawings, wherein, by way of illustration and example, the aspects of the present invention are disclosed.

BRIEF DESCRIPTION OF DRAWINGS

[0019] The present invention will be better understood after reading the following detailed description of the presently preferred aspects thereof with reference to the appended drawings, in which the features, other aspects and advantages of certain exemplary embodiments of the invention will be more apparent from the accompanying drawing in which:

[0020] Figure 1a illustrates a pictorial representation of the defect with dead cells in a solar panel;

[0021] Figure 1b illustrates a pictorial representation of the defect with crack in the solar panel; and

[0022] Figure 2 illustrates a flowchart of a system of defect detection in the solar panel.

SUMMARY OF THE INVENTION

[0023] The present invention relates to an artificial intelligence (AI)-based system for detecting defects in a solar panel. The system comprises of a solar panel, to be tested for defects, an electroluminescence (EL) testing device, a display unit wirelessly connected to the EL testing device, an imaging device, installed in proximity of the display unit, an electronic device wiredly connected to the imaging device and a server, having a database, wirelessly connected to the electronic device. The electroluminescence (EL) testing device further comprises of a power supply unit integrated in the EL testing device and an infrared imaging device. Furthermore, the electronic device is installed with a graphics processing unit. Additionally, the graphics processing unit comprises a machine learning module for executing an instruction to perform defect analysis in the images directed by an artificial intelligence module and upon successful defect analysis the solar panel sent for next manufacturing stage.

[0024] The present invention also provides a method for detecting defects in a solar panel. The method comprises steps of: placing a solar panel inside an electroluminescence (EL) testing device, shielding the device chamber from external light to improve an image clarity, supplying a power to the solar panel by connecting a positive and a negative terminal of the power supply to a solar panel busbar to activate an electroluminescent effect, capturing an image of an emitted light to form an electroluminescence (EL) image created by the electroluminescence effect in the solar panel through an infrared imaging device integrated in the electroluminescence (EL) testing device, displaying the EL image by a display unit wirelessly connected to the EL testing device, capturing images of the EL image displayed on the display unit by an imaging device, processing the captured image by a graphics processing unit installed in an electronic device wiredly connected to the imaging unit, monitoring multi-site images remotely and centrally in real time by a server wirelessly connected to the electronic device and storing the defects detection result and manual correction after analysis of the defects in the solar panel in a database.

DETAILED DESCRIPTION OF INVENTION

[0025] The following detailed description and embodiments set forth herein below are merely exemplary out of the wide variety and arrangement of instructions which can be employed with the present invention. The present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. All the features disclosed in this specification may be replaced by similar other or alternative features performing similar or same or equivalent purposes. Thus, unless expressly stated otherwise, they all are within the scope of the present invention.

[0026] Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the invention. In addition, descriptions of well-known functions and constructions are omitted for clarity and conciseness.

[0027] The terms and words used in the following description and claims are not limited to the bibliographical meanings but are merely used to enable a clear and consistent understanding of the invention. Accordingly, it should be apparent to those skilled in the art that the following description of exemplary embodiments of the present invention are provided for illustration purpose only and not for the purpose of limiting the invention.

[0028] It is to be understood that the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise.

[0029] It should be emphasized that the term “comprises/comprising” when used in this specification is taken to specify the presence of stated features, integers, steps, or components but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof.

[0030] Accordingly, the present invention relates to detect defects in a solar panel. Specifically, the present invention relates to a system of detecting defects in the solar panel using electroluminescence (EL) images and visual images in solar panel. More specifically, the present invention relates to the system that combines advanced artificial intelligence and machine learning modules with high-resolution imaging to provide accurate, efficient, and reliable defect detection, to ensure only high-quality solar panels are used in power plants, thereby increasing the efficiency and lifespan of the panels.

[0031] The present invention relates to an artificial intelligence (AI) based system (100) of detecting defects in the solar panel (102). The system (100) leverages artificial intelligence to automate the defect detection process of solar panel (102) using a solar module electroluminescence (EL) testing device (104). This approach enhances the accuracy, efficiency, and reliability of defect identification in the solar panel (102). The system (100) automating the defect detection process reduces human labor and error and enhances the throughput of the inspection process, allowing for quicker and more reliable quality control.

[0032] In a preferred embodiment of the present invention, an artificial intelligence (AI) based system (100) comprises of a solar panel (102), to be tested for defects, an electroluminescence (EL) testing device (104), a display unit (112) wirelessly connected to the EL testing device (104), a visible light imaging device (114), installed in proximity of the display unit (112), an electronic device (116) wiredly connected to the visible light imaging device (114) and a server (124), having a database (126), wirelessly connected to the electronic device (116).

[0033] Figure 1(a)-2 of the present invention illustrates the complete system of the present invention including the following components:

[0034] A) Solar panel (102): A solar panel (102) converts solar energy into electricity to be used to power homes and industries. The present invention ensures only high-quality solar panels are used in power plants, thereby increasing the efficiency and lifespan of solar panels (102). The solar panel (102) is tested for defects such that the defects are detected rectified at right time.

[0035] B) Electroluminescence testing device (104): An electroluminescence testing device (104) enables user to detect defects in the solar panel (102). The solar panel (102) is placed inside the EL testing device (104) and chambers of the EL testing device (104) is shielded from external light to improve image clarity. The EL testing device (104) comprises: a power supply unit (106) and an infrared imaging device (108).

[0036] C) Power supply unit (106): A power supply unit (106) is integrated in the EL testing device (104) and a low electric current is passed through the solar panel by connecting positive and negative terminals to the panel’s leads (busbars) to activate the electroluminescence effect.

[0037] D) Infrared imaging device (108): An infrared imaging device (108) is integrated in the electroluminescence testing device (104) captures an emitted light created by the electroluminescence effect in the solar panel (102) to form an electroluminescence (EL) image (110). The cells present in the solar panel (102) that are defective, emit less light and appear darker in the image.

[0038] In an exemplary embodiment, as shown in figure 1(a) and 1(b), the imaging device (108) captures images with the defects including micro cracks, cells that are broken completely or have damages in individual cell, issues with electrical continuity in the conductive paths, misalignment of solar cells during production, dry soldering, edge fractures, and short circuits.

[0039] E) Display unit (112): A display unit (112) is wirelessly connected to the EL testing device (104) for displaying the EL image (110). In an exemplary embodiment, the display unit (112) may be such as, but not limited, liquid crystal display unit (LCD), plasma display unit, and cathode ray tube display unit (CRT).

[0040] F) Visible Light Imaging device (114): A visible light imaging device (114) is installed in proximity of the display unit (112), for capturing an image of the EL image (110) displayed on the display unit (112). The visible light imaging device (114) is a high-resolution external camera independent of the EL testing device (104). The visible light imaging device (114) enables the system (100) to detect and display real-time defects.

[0041] G) Electronic device (116): An electronic device (116) is wiredly connected to the visible light imaging device (114) for processing the images captured by the visible light imaging device (114). The configuration of electronic device (116) with the visible light imaging device (114) ensures efficient, real-time defect detection, enabling prompt identification of defects to maintain high production quality.

[0042] Further, the electronic device (116) is installed with a graphics processing unit (118) for processing of the captured images. The graphics processing unit (118) is installed with a machine learning module (120) for executing an instruction to perform defect analysis and corrective suggestion directed by an artificial intelligence module (122) and upon successful defect analysis the solar panel (102) sent for next manufacturing stage

[0043] H) Graphics processing unit (118): A graphics processing unit (118) processes the images in an efficient manner such that there is less or no human intervention. Further, the machine learning module (120) executes an instruction to perform defect analysis and corrective suggestion directed by the artificial intelligence module (122). The solar panel (102) with defects beyond pre-set thresholds are flagged for rework or rejection.

[0044] In an exemplary embodiment, the electronic device (116) may be such as, but not limited to, laptop, desktop, tablet and the like.

[0045] In an exemplary embodiment of the present invention, the dataset of images may include, but not limited to, consisting of 800 images, with each image having a resolution of 640x640 pixels, focusing on real-time monitoring. Further, dataset expansion may include 5900 images with a resolution of 600*600 pixels including electroluminescence (EL) images and visual images. The dataset may include category of the images, not limited to, broken cells: 250 images, black spots: 175 images, black borders: 170 images, scratches: 105 images and no electricity: 100 images.

[0046] In an exemplary embodiment, the graphics processing unit (118) performs processing such as, but not limited to, super-resolution, compression and cropping, noise reduction, and edge enhancement. The machine learning module (120) is configured to utilize yolov5 model for defect analysis and corrective suggestion.

[0047] I) Server (124): A server (124) is wirelessly connected to the electronic device (116) for real time and centralized multi-site monitoring. The server (124) further comprises of a database (126) that stores defects detection result and manual correction data after analysis of the defects in the solar panel (102).

[0048] In an exemplary embodiment, the system (100) of the present invention is built using flask framework for backend services and visual studio is used as the development environment. The system (100) consists of the graphics processing unit (118) for digital image processing and scalability. The machine learning module (120) utilizes architectures like YoloV5 model to improve detection accuracy. In addition to detecting defects, the system (100) includes an artificial intelligence (AI) module (122) that provides repair instructions that suggests or even initiates repairs automatically based on the severity of the defect. The system (100) involves flagging soldering issues for further inspection, providing real-time feedback to the user.

[0049] In an exemplary embodiment, the system (100) detects defects such as, but not limited to, broken cells: irregularities in the structure of solar cells that affect their functionality, black spots: areas of high temperature that indicate inefficiency, black borders: electrical connection failures at the borders of cells, scratches: physical damage that can reduce efficiency, no electricity: cells that fail to produce electricity.

[0050] In an embodiment, the present invention also provides a method (200) for detecting defects in the solar panel (102), comprises the following steps:-
i. placing a solar panel (102) inside an electroluminescence (EL) testing device (104);
ii. shielding the device chamber from external light to improve an image clarity;
iii. supplying a power to the solar panel (102) by connecting a positive and a negative terminal of the power supply (106) to a solar panel (102) busbar to activate an electroluminescent effect;
iv. capturing an image of an emitted light to form an electroluminescence (EL) image (110) created by the electroluminescence effect in the solar panel (102) through an infrared imaging device (108) integrated in the electroluminescence (EL) testing device (104);
v. displaying the EL image (110) by a display unit (112) wirelessly connected to the EL testing device (104);
vi. capturing images of the EL image (110) displayed on the display unit (112) by a visible light imaging device (114);
vii. processing the captured image by a graphics processing unit (118) installed in an electronic device (116) wiredly connected to the imaging unit;
viii. monitoring multi-site images remotely and centrally in real time by a server (124) wirelessly connected to the electronic device (116); and
ix. storing the defects detection result and manual correction after analysis of the defects in the solar panel (102) in a database (126) installed in the server (124).

[0051] Further, the defect analysis by the artificial intelligence module (122) comprises steps of:
a) collecting data of a plurality of the captured images by the imaging device (114) with defects under different conditions;
b) applying data annotation of the collected data to get label of the images with type of defect;
c) augmenting the data for transforming the labelled images such as rotating, flipping, and changing brightness to increase dataset diversity;
d) classifying different types of defects of the augmented data of the images such as cracks, corrosion;
e) training the machine learning module (120) for analysis of the classified defects in the image;
f) providing corrective suggestion for improvement of the defects by the machine learning module (120); and
g) reviewing the suggestion and correcting the defects in the solar panel (102) as per the suggestion.

[0052] Further, as shown in figure 2, flowchart of the method comprises steps of:
o beginning the defect detection process, at step 302;
o initializing and loading the machine learning model for defect detection in images, at step 304;
o collecting images of photovoltaic cells under different conditions, at step 306;
o labelling images with defect types, preparing them for model training, at step 308;
o applying transformations like rotation, flipping, and brightness changes to increase dataset diversity, at step 310;
o classifying the defects after transformation, at step 312;
o training the machine learning model using the collected and augmented data, at step 314;
o capturing standard images of the photovoltaic cells in real-time during inspection, at step 316;
o capturing Electroluminescence (EL) images to enhance defect detection, at step 318;
o processing the captured standard images to prepare them for machine learning model inference, at step 320;
o processing the EL images similarly, optimizing them for model inference, at step 322;
o using the machine learning model to detect defects in standard images, at step 324;
o displaying result in a module, at step 326;
o detecting defects in the EL images using the machine learning model, at step 328;
o refining the detection results from both image types, consolidating insights, at step 330;
o providing potential corrective actions based on detected defects, at step 332;
o allowing users to manually review and adjust the results if necessary, at step 334;
o saving the detection results and any manual corrections in a database, at step 336;
o generating logs and comprehensive reports based on detection results, at step 338; and
o ending the defect detection and classification process, at step 340.

[0053] In an embodiment the advantages of the present invention are enlisted herein:
• The present invention develops an automated system to detect multiple types of defects in photovoltaic cells of the solar panel such as, dry soldering, latent defects and micro-cracks.
• The present invention develops an automated system to detect a wide range of defects such as broken cells, black borders, scratches and no electricity in solar cells of the solar panel through integration of electroluminescence (EL) and visual images.
• The present invention implements a real-time monitoring solution.
• The present invention integrates artificial intelligence (AI)-driven repair systems to suggest or initiate repairs.
• The present invention provides a system capable of detecting defects with a better accuracy.
• The present invention provides an artificial intelligence (AI) powered repair suggestions, reducing time and increasing production efficiency.
• The present invention provides a method that does not damage the solar panel during testing.
• The present invention identifies potential issues that cause failure in the solar panel during operation.
• The present invention detects defects at early stage in production, allowing timely rework.
• The present invention improves the robustness and reliability of defect detection by using diverse image data, ensuring a comprehensive analysis of defects in different stages of solar panel production.

[0054] While this invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims.

, Claims:WE CLAIM:
1. An Artificial Intelligence (AI)-based system (100) for detecting defects in a solar panel, comprising:
a) the electroluminescence (EL) testing device (104) configured to test the solar panel for defects, comprising:
o a power supply unit (106) integrated in the EL testing device (104) for activating an electroluminescence effect in the solar panel (102); and

o an infrared imaging device (108) integrated in the EL testing device (104) for capturing an emitted light created by the electroluminescence effect in the solar panel (102) to form an electroluminescence (EL) image (110);
b) a display unit (112) wirelessly connected to the EL testing device (104) for displaying the EL image (110);
c) a visible light imaging device (114), installed in proximity of the display unit (112), to capture images of the EL image (110) displayed on the display unit (112);
d) an electronic device (116) wiredly connected to the visible light imaging device (114); and
e) a server (124) comprising a database (126), wirelessly connected to the electronic device (116) for real time and centralized multi-site monitoring.

2. The system (100) as claimed in claim 1, wherein the electronic device (116) comprises a graphics processing unit (118) installed with a machine learning module (120) for executing an instruction to perform defect analysis and corrective suggestion directed by an artificial intelligence module (122).

3. The system (100) as claimed in claim 1, wherein the infrared imaging device is configured to capture the image of the emitted light created by an electroluminescence effect in the solar panel.

4. The system (100) as claimed in claim 1, wherein the graphics processing unit (118) assigns defect labels to each defect detected in the EL image (110).

5. The system (100) as claimed in claim 1, wherein the solar panel (102) with defects beyond pre-set thresholds are flagged for rework or rejection.

6. The system (100) as claimed in claim 1, wherein the machine learning module (120) is configured to utilize yolov5 model for defect analysis and corrective suggestion.

7. The system (100) as claimed in claim 1, wherein the database (126) stores defects detection result and manual correction data after analysis of the defects in the solar panel (102).

8. A method (200) for detecting defects in a solar panel by the system (100) as claimed in claim 1, the method comprising steps:
i. placing a solar panel (102) inside an electroluminescence (EL) testing device (104);
ii. shielding the device chamber from external light to improve an image clarity;
iii. supplying a power to the solar panel (102) by connecting a positive and a negative terminal of the power supply (106) to a solar panel (102) busbar to activate an electroluminescent effect;
iv. capturing an image of an emitted light to form an electroluminescence (EL) image (110) created by the electroluminescence effect in the solar panel (102) through an infrared imaging device (108) integrated in the electroluminescence (EL) testing device (104);
v. displaying the EL image (110) by a display unit (112) wirelessly connected to the EL testing device (104);
vi. capturing images of the EL image (110) displayed on the display unit (112) by an imaging device (114);
vii. processing the captured image by a graphics processing unit (118) installed in an electronic device (116) wiredly connected to the imaging unit;
viii. monitoring multi-site images remotely and centrally in real time by a server (124) wirelessly connected to the electronic device (116); and
ix. storing the defects detection result and manual correction after analysis of the defects in the solar panel (102) in a database (126) installed in the server (124).

9. The method as claimed in claim 8, wherein the graphics processing unit (118) comprising a machine learning module (120) executing an instruction to perform defect analysis in the images directed by an artificial intelligence module (122) and upon successful defect analysis the solar panel (102) sent for next manufacturing stage.

10. The method as claimed in claim 9, wherein the defect analysis by the artificial intelligence module (122) comprising steps:
a. collecting data of a plurality of the captured images by the imaging device (114) with defects under different conditions;
b. applying data annotation of the collected data to get label of the images with type of defect;
c. augmenting the data for transforming the labelled images such as rotating, flipping, and changing brightness to increase dataset diversity;
d. classifying different types of defects of the augmented data of the images such as cracks, corrosion;
e. training the machine learning module (120) for analysis of the classified defects in the image;
f. providing corrective suggestion for improvement of the defects by the machine learning module (120); and
g. reviewing the suggestion and correcting the defects in the solar panel (102) as per the suggestion.