Echo state network-based spatio-temporal model for solar irradiance estimation
Qian Li, Zhou Wu, Rui Ling, Mi Tan · Energy Procedia · 2019
In this study, echo state network-based spatio-temporal (STESN) model is proposed to predict the solar irradiance of a given target station, depending on the historical data of its surrounding sites. Hourly solar irradiance datasets for the whole year 2017, obtained from California Irrigation Management Information System (CIMIS) at eight different stations are used. Comparison studies show that SPESN could achieve more accurate prediction, compared with the conventional persistence (PSS) benchmark model.