Fast and non-destructive analysis of material defect is a crucial demand for semiconductor devices. Herein, we are devoted to exploring a solar-cell defect analysis method based on machine learning of the mo. Electronic defect is one of the most fundamental and important physical properties of a. 2.1. Charge-carrier mechanism of perturbation TPVIn a complete cell, charge-carrier processes are determined by a series of time-dependent charg. In this work, based on a comprehensive understanding of the generation and decay mechanism of the perturbation photovoltage, we have explored to develop a defect analysis. Y. S. Li, J. Shi and Q. Meng conceived the idea. Y. S. Li conducted device simulation, machine learning programming, data analysis and paper writing. Y. M. Li contributed to th. The authors are very grateful to Prof. Yuan Lin (Institute of Chemistry, Chinese Academy of Science), Dr. Nicola Courtier (University of Oxford, UK), and Dr. Haili Wang (COMSO.
[PDF Version]
Many existing methods for detecting solar cell defects focus on the analysis of electroluminescence (EL) infrared images, specifically in the 1000–1200 nm wave length range. Chiou et al. (2011) developed a regional growth detection algorithm to extract cracks defects from the captured images.
Surface defects in solar cells are various and can be challenging to detect due to the complex background. Before the widespread use of Convolutional Neural Networks (CNNs), manually extracting features for defect detection was a common method in machine vision. The passage discusses the difficulties of this approach.
Can deep learning detect solar cells based on a defect-free model?
The deep belief network is an unsupervised learning method that can reconstruct a defect-free model based on the current image of solar cells. However, it uses a small number of data sets. There have been no reports about surface defect detection of solar cells using deep learning.
Which ML-based techniques are used for surface defect detection of solar cells?
ML-based techniques for surface defect detection of solar cells were reviewed by Rana and Arora, of which were only imaging-based techniques. Similarly, Al-Mashhadani et al., have reviewed DL-based studies that adopted only imaging-based techniques.
Can machine learning detect solar cell surface defects?
It can be seen from the experimental results that the detection of solar cell surface defects using machine learning methods like LBP + HOG-SVM and Gabor-SVM is not very effective. The precision is 10% lower and the recall is 8% lower compared to CNN methods.
Can image-based defect detection improve solar cell surface quality control?
Image-based defect detection has been employed in the solar cell manufacturing industry for improving the production quality of the solar cell module through surface inspection. This method can also increase the lifetime of the solar cell module.