Short-term PV power forecasting system Based on GraphCast Weather Forecasting Model

With the inability to accurately predict photovoltaic (PV) output under extreme weather conditions, this paper proposes a novel method for predicting PV output based on the GraphCast weather forecasting model. By utilizing the predictive capabilities of the GraphCast model to forecast future weather data and integrating it with a deep learning model to predict PV output, simulations have demonstrated that this method offers higher accuracy, especially under extreme weather conditions.