Short-Term Load Forecasting Using Convolutional Neural Networks in COVID-19 Context: The Romanian Case Study

Short-term load forecasting (STLF) is fundamental for the proper operation of power systems, as it finds its use in various basic processes. Therefore, advanced calculation techniques are needed to obtain accurate results consumption prediction, taking into account numerous exogenous factors that influence results’ precision. The purpose this study integrate, additionally conventional (weather, holidays, etc.), current aspects regarding global COVID-19 pandemic solving STLF problem, using a convolutional neural network (CNN)-based model. To evaluate and validate impact new variables considered model, simulations conducted publicly available data from Romanian system. A comparison further carried out assess performance proposed multiple linear regression method provided by Transmission System Operator (TSO). In regard, Mean Squared Error (MSE), Absolute (MAE), Percentage (MAPE), Root Square (RMSE) used evaluation indexes. methodology shows great potential, reveal better error values compared TSO results, despite limited historical data.