Solar radiation estimation for photovoltaic power generation using Geostationary and Low-Earth Orbit Satellite Data

Solar radiation measurement data is used to predict photovoltaic power generation. Ground measurement is highly accurate, but there is a problem the measurement location is limited. On the other hand, satellite measurement is available anywhere on the earth, but the accuracy is not sufficient. Therefore, this paper has proposed to model the relationship between ground measurement data and satellite measurement data using machine learning. Focusing on the recently developed constellation technology that links multiple low earth orbit satellite to observe data with high temporal resolution, this paper has proposed a solar radiation estimation model that combines geostationary orbit satellite data and low earth orbit satellite data as input data. This paper aims to estimate the solar radiation on the ground using satellite measurement data and improve the estimation accuracy.