2026-05-11T03:55:00.637Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:9638 2024-12-03T20:10:47Z 3

Modified Teaching-Learning-Based Algorithm Tuned Long Short-Term Memory for Household Energy Consumption Forecasting, Chapter in AIS Algorithms for Intelligent Systems: WCAIAA 2023: World Conference on Artificial Intelligence: Advances and Applications, Springer 2023 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/9638 https://link.springer.com/chapter/10.1007/978-981-99-5881-8_28 L. Jovanovic M. Kljajic A. Petrovic V. Mizdrakovic M. Zivkovic N. Bacanin Energy production and conservation have become increasingly pressing issues in recent years. Resource scarcity and environmental awareness have led to an increased demand for clean renewable energy sources. However, fluctuations in power demand make managing increasingly complex power distribution systems a challenge. It is evident that reliable system for forecasting energy demand is necessary. This work proposes a novel artificial intelligence-based model for the prediction of individual household energy consumption using long short-term memory artificial neural networks. As the performance of these networks is highly reliant on appropriate control parameter selection this work proposed a novel modified teaching-learning-based optimization algorithm used to optimize the process of selection. The introduced method has been tested on a real-world energy demand dataset for individual households and subjected to a comparative analysis against several other well-established algorithms. The results indicate that the novel introduced technique outperformed competing algorithms and that it is viable for forecasting individual household energy demand. Overall results of the best LSTM model had mean absolute error of 0.024481, which confirms previous findings. bookPart Springer, Singapore 347 362 10.1007/978-981-99-5881-8_28 AIS Algorithms for Intelligent Systems: WCAIAA 2023: World Conference on Artificial Intelligence: Advances and Applications