A granular modeling method for non-uniform panel degradation based on I–V characterization and electroluminescence imaging. (October 2021)

• Record Type: Journal Article
• Authors: Garaj, Martin; Chung, Henry Shu-Hung; Spataru, Sergiu; Lo, Alan Wai-Lun; Wang, Huai
• Is Part Of: Solar energy. Volume 227(2021)
• Page Start: 162
• Page End: 178
• Publication Date: 2021-10
• Subjects: 00-01 -- 99-00; Evolutionary algorithm -- Degradation modeling -- Preventive diagnostics; Solar energy -- Periodicals; Solar engines -- Periodicals; 621.47
• DOI: 10.1016/j.solener.2021.08.069
• Languages: English
• ISSNs: 0038-092X

Abstract

Abstract: A new model of photovoltaic (PV) panel is proposed. The model precisely replicates sub-cell level degradation, such as cracks and interconnect failures, and reproduces their effect at the panel level I–V characteristic. Moreover, a regression method is proposed, which infers the model's parameters from combination of electroluminescent (EL) image and degraded I–V characteristic. The combination of quantitative (EL image) and qualitative (I–V characteristic) enables to characterize the degradation of the cells embedded in the PV panel, without physical access to the cells. The proposed model and the regression problem is experimentally verified on a set of 3 single cell measurements and a set of 4 crystalline PV panels with various levels of degradation. Highlights: Model of a PV panel incorporating non-uniform degradation. Degradation characterization from I–V characteristic and Electroluminescent image. Efficient model evaluation by Newton–Raphson method. Model parameter identification using Evolutionary Algorithm.