TY - GEN
T1 - Deep learning with LSTM networks for medium to long range demand forecasting
T2 - 31st International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2018
AU - Bouktif, Salah
AU - Fiaz, Ali
N1 - Publisher Copyright:
© 2018 University of Minho. All rights reserved.
PY - 2018
Y1 - 2018
N2 - Accurate load forecasting has become increasingly important as it can help power companies for better scheduling to reduce excessive electricity production. This study presents an approach that uses long short-term memory (LSTM) neural network to construct a forecasting model for medium to long term aggregate load forecasting. An array of linear and non-linear machine learning models are also built on electric load dataset and the best performing model is used as benchmark for comparison with deep learning approach. Weather related variables temperature, humidity and wind speed are used as exogenous inputs. The proposed methods are applied on a publicly available dataset of electric load consumption in metropolitan France. The goal is to ascertain with what accuracy can aggregated load values in Megawatts can be predicted using machine learning and deep learning approaches. Results show that LSTM based prediction model using only time lags gives low forecast errors and outperformed machine learning models optimized with hyperparameter tuning.
AB - Accurate load forecasting has become increasingly important as it can help power companies for better scheduling to reduce excessive electricity production. This study presents an approach that uses long short-term memory (LSTM) neural network to construct a forecasting model for medium to long term aggregate load forecasting. An array of linear and non-linear machine learning models are also built on electric load dataset and the best performing model is used as benchmark for comparison with deep learning approach. Weather related variables temperature, humidity and wind speed are used as exogenous inputs. The proposed methods are applied on a publicly available dataset of electric load consumption in metropolitan France. The goal is to ascertain with what accuracy can aggregated load values in Megawatts can be predicted using machine learning and deep learning approaches. Results show that LSTM based prediction model using only time lags gives low forecast errors and outperformed machine learning models optimized with hyperparameter tuning.
KW - Deep learning
KW - Long short-term memory (LSTM)
KW - Machine learning
KW - Medium to long-term load forecasting
KW - Recurrent neural network
UR - http://www.scopus.com/inward/record.url?scp=85064177043&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85064177043&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85064177043
T3 - ECOS 2018 - Proceedings of the 31st International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems
BT - ECOS 2018 - Proceedings of the 31st International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems
PB - University of Minho
Y2 - 17 June 2018 through 21 June 2018
ER -