Estimation and correction of PN junction temperature of photovoltaic modules for improving photovoltaic power prediction accuracy

The PN junction temperature (T_cell) is a crucial factor in the lifetime of photovoltaic (PV) modules, which is a significant parameter that impacts the output power of PV stations. However, the current model to estimate the T_cell of PV modules in real operating conditions suffers from low precision and limited application scenes. This paper proposes a method to estimate the T_cell by incorporating multiple environmental factors. Firstly, an optimization algorithm is employed in obtaining the T_cell by fitting the output power characteristics of PV modules under various operating conditions. Subsequently, the ambient temperature and other meteorological data are utilized as inputs in training a neural network, with the T_cell to be the output. Finally, the trained neural network is employed to calculate T_cell for the specific operating conditions. The resulting T_cell is subsequently plugged into the transformation equation to yield a more precise estimation of the PV module’s output characteristics.