Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction
Electric consumption prediction methods are investigated for many reasons, such as decision-making related to energy efficiency as well as for anticipating demand and the dynamics of the energy market. The objective of the present work is to compare two Deep Learning models, namely the Long Short-Te...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-11-01
|
| Series: | Energy Reports |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352484723014208 |
| _version_ | 1827400881232936960 |
|---|---|
| author | Davi Guimarães da Silva Anderson Alvarenga de Moura Meneses |
| author_facet | Davi Guimarães da Silva Anderson Alvarenga de Moura Meneses |
| author_sort | Davi Guimarães da Silva |
| collection | DOAJ |
| description | Electric consumption prediction methods are investigated for many reasons, such as decision-making related to energy efficiency as well as for anticipating demand and the dynamics of the energy market. The objective of the present work is to compare two Deep Learning models, namely the Long Short-Term Memory (LSTM) model, and the Bi-directional LSTM (BLSTM) for univariate electric consumption Time Series (TS) short-term forecast model. The Data Sets (DSs) were selected for their different contexts and scales, with the goal of assessing the robustness of the models. Four DSs were used, related to the power consumption of: (a) a household in France; (b) a university building in Santarém, Brazil; (c) the Tétouan city zones, in Morocco; and (d) the aggregated electric demand of Singapore. The metrics RMSE, MAE, MAPE and R2 were calculated in a TS cross-validation scheme. Friedman’s test was applied to normalized RMSE (NRMSE) results, showing that BLSTM outperforms LSTM with statistically significant difference (p = 0.0455), corroborating the fact that bidirectional weight updating significantly improves the LSTM performance with respect to different scales of electric power consumption. The present work provides statistical evidence supporting the conclusion that BLSTM outperforms LSTM models according to the tests performed, based on a complete methodology for TS prediction, and also establishes a baseline for future investigation of electric consumption TS prediction. |
| first_indexed | 2024-03-08T20:11:05Z |
| format | Article |
| id | doaj.art-5a8c20c3f8024878a7eee37c4b8b691b |
| institution | Directory Open Access Journal |
| issn | 2352-4847 |
| language | English |
| last_indexed | 2024-03-08T20:11:05Z |
| publishDate | 2023-11-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Energy Reports |
| spelling | doaj.art-5a8c20c3f8024878a7eee37c4b8b691b2023-12-23T05:21:51ZengElsevierEnergy Reports2352-48472023-11-011033153334Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption predictionDavi Guimarães da Silva0Anderson Alvarenga de Moura Meneses1Federal University of Western Pará – Graduate Program in Society, Nature and Development, R. Vera Paz, s/n, Salé, CEP 68.035-110 Santarém, PA, Brazil; Federal Institute of Education, Science and Technology of Pará, Brazil; Federal University of Western Pará, Institute of Geosciences and Engineering, Laboratory of Computational Intelligence, Brazil; Corresponding author at: Federal University of Western Pará – Graduate Program in Society, Nature and Development, R. Vera Paz, s/n, Salé, CEP 68.035-110 Santarém, PA, Brazil.Federal University of Western Pará – Graduate Program in Society, Nature and Development, R. Vera Paz, s/n, Salé, CEP 68.035-110 Santarém, PA, Brazil; Federal University of Western Pará, Institute of Geosciences and Engineering, Laboratory of Computational Intelligence, BrazilElectric consumption prediction methods are investigated for many reasons, such as decision-making related to energy efficiency as well as for anticipating demand and the dynamics of the energy market. The objective of the present work is to compare two Deep Learning models, namely the Long Short-Term Memory (LSTM) model, and the Bi-directional LSTM (BLSTM) for univariate electric consumption Time Series (TS) short-term forecast model. The Data Sets (DSs) were selected for their different contexts and scales, with the goal of assessing the robustness of the models. Four DSs were used, related to the power consumption of: (a) a household in France; (b) a university building in Santarém, Brazil; (c) the Tétouan city zones, in Morocco; and (d) the aggregated electric demand of Singapore. The metrics RMSE, MAE, MAPE and R2 were calculated in a TS cross-validation scheme. Friedman’s test was applied to normalized RMSE (NRMSE) results, showing that BLSTM outperforms LSTM with statistically significant difference (p = 0.0455), corroborating the fact that bidirectional weight updating significantly improves the LSTM performance with respect to different scales of electric power consumption. The present work provides statistical evidence supporting the conclusion that BLSTM outperforms LSTM models according to the tests performed, based on a complete methodology for TS prediction, and also establishes a baseline for future investigation of electric consumption TS prediction.http://www.sciencedirect.com/science/article/pii/S2352484723014208Electric consumption forecastDeep learningUnivariate time seriesDeep neural networksLong-Short Term Memory |
| spellingShingle | Davi Guimarães da Silva Anderson Alvarenga de Moura Meneses Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction Energy Reports Electric consumption forecast Deep learning Univariate time series Deep neural networks Long-Short Term Memory |
| title | Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction |
| title_full | Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction |
| title_fullStr | Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction |
| title_full_unstemmed | Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction |
| title_short | Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction |
| title_sort | comparing long short term memory lstm and bidirectional lstm deep neural networks for power consumption prediction |
| topic | Electric consumption forecast Deep learning Univariate time series Deep neural networks Long-Short Term Memory |
| url | http://www.sciencedirect.com/science/article/pii/S2352484723014208 |
| work_keys_str_mv | AT daviguimaraesdasilva comparinglongshorttermmemorylstmandbidirectionallstmdeepneuralnetworksforpowerconsumptionprediction AT andersonalvarengademourameneses comparinglongshorttermmemorylstmandbidirectionallstmdeepneuralnetworksforpowerconsumptionprediction |