Hassan, Sharmarke and Dhimish, Mahmoud (2023) Dual spin max pooling convolutional neural network for solar cell crack detection. Scientific Reports. 11099. ISSN 2045-2322
Abstract
This paper presents a solar cell crack detection system for use in photovoltaic (PV) assembly units. The system utilizes four different Convolutional Neural Network (CNN) architectures with varying validation accuracy to detect cracks, microcracks, Potential Induced Degradations (PIDs), and shaded areas. The system examines the electroluminescence (EL) image of a solar cell and determines its acceptance or rejection status based on the presence and size of the crack. The proposed system was tested on various solar cells and achieved a high degree of accuracy, with an acceptance rate of up to 99.5%. The system was validated with thermal testing using real-world cases, such as shaded areas and microcracks, which were accurately predicted by the system. The results show that the proposed system is a valuable tool for evaluating the condition of PV cells and can lead to improved efficiency. The study also shows that the proposed CNN model outperforms previous studies and can have significant implications for the PV industry by reducing the number of defective cells and improving the overall efficiency of PV assembly units.
Metadata
Item Type: | Article |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © The Author(s) 2023 |
Dates: |
|
Institution: | The University of York |
Academic Units: | The University of York > Faculty of Sciences (York) > Electronic Engineering (York) |
Depositing User: | Pure (York) |
Date Deposited: | 12 Jul 2023 23:11 |
Last Modified: | 21 Dec 2024 00:23 |
Published Version: | https://doi.org/10.1038/s41598-023-38177-8 |
Status: | Published |
Refereed: | Yes |
Identification Number: | 10.1038/s41598-023-38177-8 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:201346 |