Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Seyed Azad Nabavi
  • Sahar Mohammadi
  • Naser Hossein Motlagh
  • Sasu Tarkoma
  • Philipp Geyer

External Research Organisations

  • University of Helsinki
View graph of relations

Details

Original languageEnglish
Pages (from-to)2873-2900
Number of pages28
JournalEnergy Reports
Volume12
Early online date5 Sept 2024
Publication statusE-pub ahead of print - 5 Sept 2024

Abstract

Forecasting electricity load plays a vital role in the planning and management of sustainable power systems, considering the multifaceted impacts of social, economic, technical, environmental, and cultural factors on electricity consumption. Addressing this complexity requires the development of robust models capable of handling high levels of nonlinearity. In this study, we used four machine learning-based methods for forecasting short to long-term electricity load. Electricity load Data were collected from the Iranian Grid Management Company (IGMC) online electricity data and German electricity market data platform. Environmental factors (ambient temperature, cloud cover, solar radiation, precipitation), social events (vacations, festivals), and time series features (Hour Lag, Day Lag, Week Lag, Year Lag) were considered as input variables. The methods include Long Short-Term Memory (LSTM), a combination of LSTM and Discrete Wavelet Transformation (DWT-LSTM), Nonlinear Auto-Regressive with eXogenous inputs (NARX), and Support Vector Machine (SVM) regressor. We apply these methods to forecast electricity load under normal conditions and during social events in both Iran and Germany, evaluating their performance using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Our results demonstrate that the DWT-LSTM method achieves the highest accuracy, with MAPE ranging from 0.59% to 4.2% for Iran and 0.29% to 3.02% for Germany, across hour-ahead to year-ahead forecasts. Moreover, during special events and festivals, DWT-LSTM exhibits precise forecasting capabilities, with MAPE ranging from 0.55% to 3.07% for Iran and 0.33% to 6.01% for Germany, spanning hour-ahead to week-ahead predictions. Comparative analysis of the implemented methods confirms the superior accuracy of DWT-LSTM, followed by LSTM, NARX, and SVM methods, respectively. Our proposed forecasting approach demonstrates high performance in anticipating electricity load under both standard conditions and during significant social events in diverse geographical contexts.

Keywords

    Deep learning, Discrete wavelet transformation, Electricity load, LSTM, LTLF, Neural network, STLF

ASJC Scopus subject areas

Sustainable Development Goals

Cite this

Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM. / Nabavi, Seyed Azad; Mohammadi, Sahar; Motlagh, Naser Hossein et al.
In: Energy Reports, Vol. 12, 12.2024, p. 2873-2900.

Research output: Contribution to journalArticleResearchpeer review

Nabavi, SA, Mohammadi, S, Motlagh, NH, Tarkoma, S & Geyer, P 2024, 'Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM', Energy Reports, vol. 12, pp. 2873-2900. https://doi.org/10.1016/j.egyr.2024.08.070
Nabavi, S. A., Mohammadi, S., Motlagh, N. H., Tarkoma, S., & Geyer, P. (2024). Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM. Energy Reports, 12, 2873-2900. Advance online publication. https://doi.org/10.1016/j.egyr.2024.08.070
Nabavi SA, Mohammadi S, Motlagh NH, Tarkoma S, Geyer P. Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM. Energy Reports. 2024 Dec;12:2873-2900. Epub 2024 Sept 5. doi: 10.1016/j.egyr.2024.08.070
Nabavi, Seyed Azad ; Mohammadi, Sahar ; Motlagh, Naser Hossein et al. / Deep learning modeling in electricity load forecasting : Improved accuracy by combining DWT and LSTM. In: Energy Reports. 2024 ; Vol. 12. pp. 2873-2900.
Download
@article{53c6fd965d584545b40be98bf08f9182,
title = "Deep learning modeling in electricity load forecasting: Improved accuracy by combining DWT and LSTM",
abstract = "Forecasting electricity load plays a vital role in the planning and management of sustainable power systems, considering the multifaceted impacts of social, economic, technical, environmental, and cultural factors on electricity consumption. Addressing this complexity requires the development of robust models capable of handling high levels of nonlinearity. In this study, we used four machine learning-based methods for forecasting short to long-term electricity load. Electricity load Data were collected from the Iranian Grid Management Company (IGMC) online electricity data and German electricity market data platform. Environmental factors (ambient temperature, cloud cover, solar radiation, precipitation), social events (vacations, festivals), and time series features (Hour Lag, Day Lag, Week Lag, Year Lag) were considered as input variables. The methods include Long Short-Term Memory (LSTM), a combination of LSTM and Discrete Wavelet Transformation (DWT-LSTM), Nonlinear Auto-Regressive with eXogenous inputs (NARX), and Support Vector Machine (SVM) regressor. We apply these methods to forecast electricity load under normal conditions and during social events in both Iran and Germany, evaluating their performance using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Our results demonstrate that the DWT-LSTM method achieves the highest accuracy, with MAPE ranging from 0.59% to 4.2% for Iran and 0.29% to 3.02% for Germany, across hour-ahead to year-ahead forecasts. Moreover, during special events and festivals, DWT-LSTM exhibits precise forecasting capabilities, with MAPE ranging from 0.55% to 3.07% for Iran and 0.33% to 6.01% for Germany, spanning hour-ahead to week-ahead predictions. Comparative analysis of the implemented methods confirms the superior accuracy of DWT-LSTM, followed by LSTM, NARX, and SVM methods, respectively. Our proposed forecasting approach demonstrates high performance in anticipating electricity load under both standard conditions and during significant social events in diverse geographical contexts.",
keywords = "Deep learning, Discrete wavelet transformation, Electricity load, LSTM, LTLF, Neural network, STLF",
author = "Nabavi, {Seyed Azad} and Sahar Mohammadi and Motlagh, {Naser Hossein} and Sasu Tarkoma and Philipp Geyer",
note = "Publisher Copyright: {\textcopyright} 2024 The Author(s)",
year = "2024",
month = sep,
day = "5",
doi = "10.1016/j.egyr.2024.08.070",
language = "English",
volume = "12",
pages = "2873--2900",

}

Download

TY - JOUR

T1 - Deep learning modeling in electricity load forecasting

T2 - Improved accuracy by combining DWT and LSTM

AU - Nabavi, Seyed Azad

AU - Mohammadi, Sahar

AU - Motlagh, Naser Hossein

AU - Tarkoma, Sasu

AU - Geyer, Philipp

N1 - Publisher Copyright: © 2024 The Author(s)

PY - 2024/9/5

Y1 - 2024/9/5

N2 - Forecasting electricity load plays a vital role in the planning and management of sustainable power systems, considering the multifaceted impacts of social, economic, technical, environmental, and cultural factors on electricity consumption. Addressing this complexity requires the development of robust models capable of handling high levels of nonlinearity. In this study, we used four machine learning-based methods for forecasting short to long-term electricity load. Electricity load Data were collected from the Iranian Grid Management Company (IGMC) online electricity data and German electricity market data platform. Environmental factors (ambient temperature, cloud cover, solar radiation, precipitation), social events (vacations, festivals), and time series features (Hour Lag, Day Lag, Week Lag, Year Lag) were considered as input variables. The methods include Long Short-Term Memory (LSTM), a combination of LSTM and Discrete Wavelet Transformation (DWT-LSTM), Nonlinear Auto-Regressive with eXogenous inputs (NARX), and Support Vector Machine (SVM) regressor. We apply these methods to forecast electricity load under normal conditions and during social events in both Iran and Germany, evaluating their performance using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Our results demonstrate that the DWT-LSTM method achieves the highest accuracy, with MAPE ranging from 0.59% to 4.2% for Iran and 0.29% to 3.02% for Germany, across hour-ahead to year-ahead forecasts. Moreover, during special events and festivals, DWT-LSTM exhibits precise forecasting capabilities, with MAPE ranging from 0.55% to 3.07% for Iran and 0.33% to 6.01% for Germany, spanning hour-ahead to week-ahead predictions. Comparative analysis of the implemented methods confirms the superior accuracy of DWT-LSTM, followed by LSTM, NARX, and SVM methods, respectively. Our proposed forecasting approach demonstrates high performance in anticipating electricity load under both standard conditions and during significant social events in diverse geographical contexts.

AB - Forecasting electricity load plays a vital role in the planning and management of sustainable power systems, considering the multifaceted impacts of social, economic, technical, environmental, and cultural factors on electricity consumption. Addressing this complexity requires the development of robust models capable of handling high levels of nonlinearity. In this study, we used four machine learning-based methods for forecasting short to long-term electricity load. Electricity load Data were collected from the Iranian Grid Management Company (IGMC) online electricity data and German electricity market data platform. Environmental factors (ambient temperature, cloud cover, solar radiation, precipitation), social events (vacations, festivals), and time series features (Hour Lag, Day Lag, Week Lag, Year Lag) were considered as input variables. The methods include Long Short-Term Memory (LSTM), a combination of LSTM and Discrete Wavelet Transformation (DWT-LSTM), Nonlinear Auto-Regressive with eXogenous inputs (NARX), and Support Vector Machine (SVM) regressor. We apply these methods to forecast electricity load under normal conditions and during social events in both Iran and Germany, evaluating their performance using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). Our results demonstrate that the DWT-LSTM method achieves the highest accuracy, with MAPE ranging from 0.59% to 4.2% for Iran and 0.29% to 3.02% for Germany, across hour-ahead to year-ahead forecasts. Moreover, during special events and festivals, DWT-LSTM exhibits precise forecasting capabilities, with MAPE ranging from 0.55% to 3.07% for Iran and 0.33% to 6.01% for Germany, spanning hour-ahead to week-ahead predictions. Comparative analysis of the implemented methods confirms the superior accuracy of DWT-LSTM, followed by LSTM, NARX, and SVM methods, respectively. Our proposed forecasting approach demonstrates high performance in anticipating electricity load under both standard conditions and during significant social events in diverse geographical contexts.

KW - Deep learning

KW - Discrete wavelet transformation

KW - Electricity load

KW - LSTM

KW - LTLF

KW - Neural network

KW - STLF

UR - http://www.scopus.com/inward/record.url?scp=85202993074&partnerID=8YFLogxK

U2 - 10.1016/j.egyr.2024.08.070

DO - 10.1016/j.egyr.2024.08.070

M3 - Article

AN - SCOPUS:85202993074

VL - 12

SP - 2873

EP - 2900

JO - Energy Reports

JF - Energy Reports

ER -