Librería Samer Atenea
Kálamo Books
Librería Elías (Asturias)
Librería Kolima (Madrid)
Librería Proteo (Málaga)
This book is devoted to the research and development of machine learning methods for forecasting and optimizing energy consumption at industrial facilities. The study is based on real-world data on daily electricity consumption at Qarmet mines for the period 2014-2025. A comparative analysis of RNN, LSTM, GRU, and Transformer models was conducted for short-term time series forecasting. Based on the experimental results, the Transformer model with a 30-day observation window and forward multi-step forecasting was found to be the most effective. Based on the forecasts, a method was developed to smooth peak loads while maintaining the total volume of energy consumption. A software package was implemented to provide forecasting, optimization, and visualization of results. The proposed approach can be used as a decision-support tool to improve energy efficiency and plan energy consumption regimes for industrial enterprises.