Forecasting I-V characteristic of PV modules considering real operating conditions using numerical method and deep learning

The current-voltage (I-V) characteristic plays a dominant role in operating a photovoltaic (PV) system as it provides information about the performance of the system. Since the output quality of PV depends mainly on the solar irradiation and cell temperature, modeling the I-V relationship regarding solar irradiation and cell temperature need to be addressed. In this paper, the long short-term memory (LSTM) model is adopted to forecast the solar irradiation and temperature of a PV module. After that, a PV module model called one-diode model is introduced to identify the I-V characteristic of the PV module, which only employs the data forecasted by the LSTM-based model and the manufactured data. Since this method combines the strengths of two techniques, it solves the uncertainty of meteorological data as well as provides an effective method to model the I-V output quality of the PV module.