Univariate solar radiation forecasting with N-BEATS

Solar radiation forecasting plays a crucial role not only in meteorology but also in various fields such as energy management and agriculture.This study aims to apply the N-BEATS model, which demonstrates excellent performance in univariate time series forecasting, to solar radiation prediction, and compare its performance with the SARIMA and LSTM models.The study first evaluates the performance of each model through simulations and then compares the prediction results using actual solar radiation data.The results from both the simulations and real-world data analysis show that the N-BEATS model outperforms the other models in terms of prediction accuracy, with significantly faster fitting speeds compared to the LSTM model.These findings suggest that the N-BEATS model can be effectively applied to various univariate time series forecasting tasks.This study demonstrates the potential of the N-BEATS model as a valuable tool for meteorological prediction tasks.