Improving Energy Efficiency Through Machine Learning-Based Load Forecasting in power systems

Energy efficiency in buildings is critical for sustainable development. The design and expansion of power systems rely heavily on electrical demand forecasts. This study explores the use of LSTM models for enhancing energy load forecasting using a large dataset covering the period from September 2021 to August 2022. The methodology involves data preprocessing techniques such as normalisation, feature selection, and outlier removal to prepare the data for model training. The LSTM model is then applied to predict energy consumption, and its performance is evaluated using various metrics, including MSE, R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>, RMSE, and MAE. The LSTM model achieves an MSE of 0.00295, R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.86627, RMSE of 0.05436, and MAE of 0.04305, demonstrating high accuracy in predicting energy loads with minimal error. A comparative analysis with other models, such as GRU and CNN-ESN, shows that LSTM outperforms these models significantly, with lower MSE, RMSE, and MAE values. The study contributes to improved energy management and operational efficiency by providing a more reliable and precise forecasting model.