Author:
Hassan Hend G.,Shahin Ahmed A.,Ziedan Ibrahim E.
Abstract
AbstractThis study predicts future values of energy consumption demand from a novel dataset that includes the energy consumption during COVID-19 lockdown, using up-to-date deep learning algorithms to reduce peer-to-peer energy system losses and congestion. Three learning algorithms, namely Random Forest (RF), Bi-LSTM, and GRU, were used to predict the future values of a building’s energy consumption. The results were compared using the RMSE and MAE evaluation metrics. The results show that predicting the future energy demand with accurate results is achievable, and that Bi-LSTM and GRU perform better, especially when trained as univariate models with only the energy consumption values and no other features included.
Publisher
Springer Science and Business Media LLC
Subject
General Earth and Planetary Sciences,General Physics and Astronomy,General Engineering,General Environmental Science,General Materials Science,General Chemical Engineering
Reference43 articles.
1. Amir V, Jadid S, Ehsan M (2019) Operation of networked multi-carrier microgrid considering demand response. COMPEL- Int J Comput Math Electr Electron Eng
2. Bishnoi D, Chaturvedi H (2021) Emerging trends in smart grid energy management systems. Int J Renew Energy Res (IJRER) 11(3):952–966
3. Bisong E et al (2019) Building machine learning and deep learning models on Google cloud platform. Springer, Cham
4. Brassington G (2017) Mean absolute error and root mean square error: which is the better metric for assessing model performance? In: EGU general assembly conference abstracts, p 3574
5. Choi Y, Park B, Jiyeon H et al (2022) Development of occupancy prediction model and performance comparison according to recurrent neural network model. Proc Architect Inst Korea 38(10):231–240