23 August 2017 Image analysis of PV module electroluminescence
Author Affiliations +
Electroluminescence imaging can be used as a non-invasive method to spatially assess performance degradation in photovoltaic (PV) modules. Cells, or regions of cells, that do not produce an infra-red luminescence signal under electrical excitation indicate potential damage in the module. In this study, an Andor iKon-M camera and an image acquisition tool provided by Andor have been utilized to obtain electroluminescent images of a full-sized multicrystalline PV module at regular intervals throughout an accelerated lifecycle test (ALC) performed in a large-scale environmental degradation chamber. Computer aided digital image analysis methods were then used to automate degradation assessment in the modules. Initial preprocessing of the images was designed to remove both background noise and barrel distortion in the image data. Image areas were then mapped so that changes in luminescent intensity across both individual cells and the full module could be identified. Two primary techniques for image analysis were subsequently investigated. In the first case, pixel intensity distributions were evaluated over each individual PV cell and changes to the intensities of the cells over the course of an ALC test were evaluated. In the second approach, intensity line scans of each of the cells in a PV module were performed and variations in line scan data were identified during the module ALC test. In this report, both the image acquisition and preprocessing technique and the contribution of each image analysis approach to an assessment of degradation behavior will be discussed.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
T. Lai, T. Lai, C. Ramirez, C. Ramirez, B. G. Potter, B. G. Potter, K. Simmons-Potter, K. Simmons-Potter, } "Image analysis of PV module electroluminescence", Proc. SPIE 10370, Reliability of Photovoltaic Cells, Modules, Components, and Systems X, 103700K (23 August 2017); doi: 10.1117/12.2274184; https://doi.org/10.1117/12.2274184

Back to Top