Wind and Solar Forecasting for Renewable Energy System using SARIMA-based Model

In order to completely fulfill a datacenter power demand, one important issue is to determine and investigate a reasonable sizing for Hybrid Renewable Energy System (HRES). Usually, in the context of datacenter renewable power supply, the energy production is hybrid and it consists of wind and solar energy production associated with battery and hydrogen energy. To design the electrical energy system, one needs to forecast weather conditions (solar radiation, wind speed) in order to evaluate the energy production yearly. The aim of this paper is to propose a SARIMA-based model for a particular renewable energy system. Indeed, thanks to the wind turbine and solar panel models, it is possible to optimize the overall cost of the global energy system. We finally validate the proposed model on actual data.