An artificial neural network-based performance model of triple-junction InGaP/InGaAs/Ge cells for the production estimation of concentrated photovoltaic systems

Analytical and empirical models analyze complex non-linear interactions between the input–output parameters of system. This is very important in case photovoltaic systems to understand their real performance potential. On other hand, manufacturers panels rate maximum system under fixed lab conditions as per standard testing (STCs) or nominal operating cell temperature (NOCT) standards IEC. These ratings do not provide actual production potential a field with fluctuating irradiance temperature. For concentrated (CPV) system, utilizing multi-junction solar cells (MJCs), there no commercial tool available production, despite some recent that also require post-processing experimental data be used conventional models. In this study, an artificial neural network (ANN)-based model presented for cell, which only convenient apply but can easily expanded predict real-field CPV any designed size. addition, ANN-based showed high accuracy 99.9% predicting output MJCs compared diode-based literature. The concentration at area are taken inputs network. If both these known, then efficiency accurately operation.