Electroluminescence Of Silicon Solar Cells And Defects Identification Based On Matlab Image Processing

Currently, energy and environmental problems make the outstanding advantages of photovoltaic more evidently. Solar cell is the core of photovoltaic system, which directly affects the quality and efficiency of PV modules. While the development of thin slices of crystalline silicon solar cells with the module manufacturing process cause many defects, such as cracks, off-grid, non-uniform resistance, flower slice and so on. Defects will greatly reduce the efficiency, reliability and service life as well as the stability of PV systems.Some of these defects are hidden or difficult to accurately determine by eyes, especially for polycrystalline silicon solar cells with grain boundary in the surface. Today actual production process is measured all by human eyes, reducing the efficiency and accuracy of detection, so a fast, efficient and accurate identification or assessment of defects in solar cells is very valuable.This paper includes three main parts:(1) Based on electroluminescence(EL)theory, by using of infrared detection method, hidden defects of crystalline silicon solar cells were detected by CCD near-infrared camera, such as cracks, off-grid, non-uniform resistance, flower slice. Then EL images were compared with visible images. It indicates that EL is an effective method for defects detection.(2) I-V characteristics of defective solar cells'were tested, It showed that fill factor, efficiency and performance went down respectively, which proves the accuracy of EL images results.(3) Dealt the EL images of defects cells with Matlab digital image processing. It identified the cracks and black cells defects.(a)The area, perimeter, and circularity of cracks were calculated, implemented to measure the cracks precisely.(b)Solar modules are processed. After rotation, single cell division and blocks identification, the area of black in single cell was calculated, the location of unqualified cells were ticked out. And a GUI graphical user interface was produced, It made the module identification fully automatic, easy for operation, met the demand of actual production operations.The contributions of the paper is making the recognition of solar cells defects automatic and precise, compared to eyes identify vaguely in nowadays'PV manufacturing industry. The automatic detection will have a large application.