Forecasting Insolation Shaded by Solar Panels for Optimal Layout in Agrivoltaic System

Yusuke Sasaki, Tamae Itabashi, Takeo Hamada, Noboru Koshizuka

Abstract: Agrivoltaics which install solar panels on farmland, have been expanding due to land competition in the global expansion of solar power generation. Insolation is an essential parameter for crop growth so it is crucial to forecast insolation under solar panels. However, previous examples of insolation estimation do not consider shading differences due to panel layout. In this paper, we constructed a machine-learning model with panel layout features to forecast the insolation shaded by solar panels. As a simulation result, the accuracy was greatly improved compared to the current method using coefficients. This method can realize the optimization of panel placement and contribute to the expansion of agrivoltaics.