Day-Ahead Load Forecasting Using Supervised Deep Learning Techniques with Day-Aware Data Arrangement: A Case Study of SUT Campus

This study proposes supervised deep learning (DL) techniques via long short-term memory (LSTM), gate recurrent unit (GRU), Bi-LSTM, and Bi-GRU to forecast electrical loads based on Day-Aware Data Arrangement. The data features are historical electrical loads with arrangement in the form of time series for the model training and evaluation. The proposed model was verified by three accuracy indices: Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Square Error (RMSE). Among the study models, the GRU model demonstrates fastest processing speed due to its simpler structure, while LSTM model has the highest forecasting efficiency.