Caracterização da resistência série em dispositivos fotovoltaicos através de imagem de eletroluminescência residual

Abstract

This dissertation proposes and validates a methodology to characterize the series resistance (Rs ) of photovoltaic devices (PVDs) from residual electroluminescence (EL) images obtained with a modified DSLR camera. The experimental setup employed a Canon EOS Rebel SL3 with the IR-cut and color-correction filters removed, allowing the capture of radiation in the visible and near-infrared spectra. Images were acquired in a dark chamber, with focus optimized by the Tenengrad metric and ISO/exposure parameters adjusted to maximize the signal-to-noise ratio. The images acquired at 10% Isc and Isc were pre-processed and used to estimate the global series resistance of the module (RSmod). As references, two conventional methods were applied: one based on manufacturer datasheet parameters and another using the slope of the I–V curve at Voc. The RSmod value obtained from EL (5.204 Ω) differed by less than 3% from the datasheet value (5.076 Ω), while the I–V slope method yielded 7.740 Ω. Based on these values, spatial Rs maps were generated, showing high structural similarity (SSIM ≈ 0.9998) and global reconstruction error below 11%. Statistical analysis identified defects corresponding to 5.6% of the active area. The proposed technique proves to be efficient and accessible, enabling accurate quantification and mapping of series resistance with potential applications in photovoltaic device diagnostics and inspection.