Multi-Step Ahead Forecasting of the Energy Consumed by the Residential and Commercial Sectors in the United States Based on a Hybrid CNN-BiLSTM Model
Abstract
:1. Introduction
2. Methodology
2.1. Spatial Features Extraction of Time Series Based on Convolutional Neural Network
2.2. Temporal Features Extraction of Time Series Based on Bidirectional Long Short-Term Memory (BiLSTM)
2.3. The CNN-BiLSTM
2.4. Forecasting Strategy for Univariate Time Series
Algorithm 1: Algorithm of recursive multi-step forecasting strategy. |
3. Hyperparameters Tuning Based on the Grid Search Approach
Algorithm 2: Grid search for tuning the optimal hyperparameters of the CNN-BiLSTM. |
4. Case Studies
4.1. Data Collection and Pre-Processing
4.2. Benchmarked Models for Comparisons and Evaluation Metrics
4.3. Forecasting Results Analysis
4.3.1. Case I: Total Energy Consumed by the Residential Sector
4.3.2. Case II: End-Use Energy Consumed by the Residential Sector
4.3.3. Case III: Primary Energy Consumed by the Commercial Sector
4.3.4. Case IV: End-Use Energy Consumed by the Commercial Sector
5. Discussion
5.1. Comparisons of the CNN-BiLSTM Model and Benchmarked Machine Learning Models
5.2. Comparisons of the CNN-BiLSTM Hybrid Model and Benchmarked Deep Learning Models
5.3. Performance Analysis of the Models Considering the COVID-19 Stay-at-Home Orders
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- World Health Organization. WHO COVID-19 Dashboard; World Health Organization: Geneva, Switzerland, 2020; Available online: https://covid19.who.int/ (accessed on 23 December 2022).
- U.S. Energy Consumption Fell by a Record 7% in 2020. Available online: https://www.eia.gov/todayinenergy/detail.php?id=47397 (accessed on 5 April 2021).
- Moreland, A.; Herlihy, C.; Tynan, M.A.; Sunshine, G.; McCord, R.F.; Hilton, C.; Poovey, J.; Werner, A.K.; Jones, C.D.; Fulmer, E.B.; et al. Timing of state and territorial COVID-19 stay-at-home orders and changes in population movement—United States, March 1–May 31, 2020. Morb. Mortal. Wkly. Rep. 2020, 69, 1198. [Google Scholar] [CrossRef]
- Stay-at-Home Orders Led to Less Commercial and Industrial Electricity Use in April. Available online: https://www.eia.gov/todayinenergy/detail.php?id=44276# (accessed on 30 June 2020).
- Chen, C.f.; de Rubens, G.Z.; Xu, X.; Li, J. Coronavirus comes home? Energy use, home energy management, and the social-psychological factors of COVID-19. Energy Res. Soc. Sci. 2020, 68, 101688. [Google Scholar] [CrossRef]
- Krarti, M.; Aldubyan, M. Review analysis of COVID-19 impact on electricity demand for residential buildings. Renew. Sustain. Energy Rev. 2021, 143, 110888. [Google Scholar] [CrossRef] [PubMed]
- Chinthavali, S.; Tansakul, V.; Lee, S.; Whitehead, M.; Tabassum, A.; Bhandari, M.; Munk, J.; Zandi, H.; Buckberry, H.; Kuruganti, T.; et al. COVID-19 pandemic ramifications on residential Smart homes energy use load profiles. Energy Build. 2022, 259, 111847. [Google Scholar] [CrossRef] [PubMed]
- Tamba, J.G.; Essiane, S.N.; Sapnken, E.F.; Koffi, F.D.; Nsouandélé, J.L.; Soldo, B.; Njomo, D. Forecasting natural gas: A literature survey. Int. J. Energy Econ. Policy 2018, 8, 216. [Google Scholar]
- Wei, N.; Li, C.; Peng, X.; Zeng, F.; Lu, X. Conventional models and artificial intelligence-based models for energy consumption forecasting: A review. J. Pet. Sci. Eng. 2019, 181, 106187. [Google Scholar] [CrossRef]
- Wright, C.; Chan, C.W.; Laforge, P. Towards developing a decision support system for electricity load forecast. In Decision Support Systems; IntechOpen: London, UK, 2012. [Google Scholar]
- Soldo, B. Forecasting natural gas consumption. Appl. Energy 2012, 92, 26–37. [Google Scholar] [CrossRef]
- Taşpınar, F.; Çelebi, N.; Tutkun, N. Forecasting of daily natural gas consumption on regional basis in Turkey using various computational methods. Energy Build. 2013, 56, 23–31. [Google Scholar] [CrossRef]
- Potočnik, P.; Soldo, B.; Šimunović, G.; Šarić, T.; Jeromen, A.; Govekar, E. Comparison of static and adaptive models for short-term residential natural gas forecasting in Croatia. Appl. Energy 2014, 129, 94–103. [Google Scholar] [CrossRef]
- He, Y.; Lin, B. Forecasting China’s total energy demand and its structure using ADL-MIDAS model. Energy 2018, 151, 420–429. [Google Scholar] [CrossRef]
- Jain, R.; Mahajan, V. Load forecasting and risk assessment for energy market with renewable based distributed generation. Renewable Energy Focus 2022, 42, 190–205. [Google Scholar] [CrossRef]
- Rakpho, P.; Yamaka, W. The forecasting power of economic policy uncertainty for energy demand and supply. Energy Reports 2021, 7, 338–343. [Google Scholar] [CrossRef]
- Zhang, Y.; Guo, H.; Sun, M.; Liu, S.; Forrest, J. A novel grey Lotka–Volterra model driven by the mechanism of competition and cooperation for energy consumption forecasting. Energy 2023, 264, 126154. [Google Scholar] [CrossRef]
- Şahin, U. Forecasting share of renewables in primary energy consumption and CO2 emissions of China and the United States under Covid-19 pandemic using a novel fractional nonlinear grey model. Expert Syst. Appl. 2022, 209, 118429. [Google Scholar] [CrossRef]
- Khan, A.M.; Osińska, M. Comparing forecasting accuracy of selected grey and time series models based on energy consumption in Brazil and India. Expert Syst. Appl. 2023, 212, 118840. [Google Scholar] [CrossRef]
- Chen, H.; Yang, Z.; Peng, C.; Qi, K. Regional energy forecasting and risk assessment for energy security: New evidence from the Yangtze River Delta region in China. J. Clean. Prod. 2022, 361, 132235. [Google Scholar] [CrossRef]
- Ding, S.; Tao, Z.; Li, R.; Qin, X. A novel seasonal adaptive grey model with the data-restacking technique for monthly renewable energy consumption forecasting. Expert Syst. Appl. 2022, 208, 118115. [Google Scholar] [CrossRef]
- Ye, L.; Dang, Y.; Fang, L.; Wang, J. A nonlinear interactive grey multivariable model based on dynamic compensation for forecasting the economy-energy-environment system. Appl. Energy 2023, 331, 120189. [Google Scholar] [CrossRef]
- Xie, N.; Liu, S. Discrete GM (1, 1) and mechanism of grey forecasting model. Syst.-Eng.-Theory Pract. 2005, 25, 93–99. [Google Scholar]
- Ahmad, T.; Zhang, D.; Huang, C. Methodological framework for short-and medium-term energy, solar and wind power forecasting with stochastic-based machine learning approach to monetary and energy policy applications. Energy 2021, 231, 120911. [Google Scholar] [CrossRef]
- Mehmood, F.; Ghani, M.U.; Ghafoor, H.; Shahzadi, R.; Asim, M.N.; Mahmood, W. EGD-SNet: A computational search engine for predicting an end-to-end machine learning pipeline for Energy Generation & Demand Forecasting. Appl. Energy 2022, 324, 119754. [Google Scholar] [CrossRef]
- Aras, S.; Hanifi Van, M. An interpretable forecasting framework for energy consumption and CO2 emissions. Appl. Energy 2022, 328, 120163. [Google Scholar] [CrossRef]
- Feng, Z.; Zhang, M.; Wei, N.; Zhao, J.; Zhang, T.; He, X. An office building energy consumption forecasting model with dynamically combined residual error correction based on the optimal model. Energy Rep. 2022, 8, 12442–12455. [Google Scholar] [CrossRef]
- Rao, C.; Zhang, Y.; Wen, J.; Xiao, X.; Goh, M. Energy demand forecasting in China: A support vector regression-compositional data second exponential smoothing model. Energy 2023, 263, 125955. [Google Scholar] [CrossRef]
- Zhu, L.; Li, M.; Wu, Q.; Jiang, L. Short-term natural gas demand prediction based on support vector regression with false neighbours filtered. Energy 2015, 80, 428–436. [Google Scholar] [CrossRef]
- Liu, H.; Tang, Y.; Pu, Y.; Mei, F.; Sidorov, D. Short-term Load Forecasting of Multi-Energy in Integrated Energy System Based on Multivariate Phase Space Reconstruction and Support Vector Regression Mode. Electr. Power Syst. Res. 2022, 210, 108066. [Google Scholar] [CrossRef]
- Jamei, M.; Ali, M.; Karbasi, M.; Xiang, Y.; Ahmadianfar, I.; Yaseen, Z.M. Designing a Multi-Stage Expert System for daily ocean wave energy forecasting: A multivariate data decomposition-based approach. Appl. Energy 2022, 326, 119925. [Google Scholar] [CrossRef]
- Li, R.; Song, X. A multi-scale model with feature recognition for the use of energy futures price forecasting. Expert Syst. Appl. 2023, 211, 118622. [Google Scholar] [CrossRef]
- Mustaqeem; Ishaq, M.; Kwon, S. A CNN-Assisted deep echo state network using multiple Time-Scale dynamic learning reservoirs for generating Short-Term solar energy forecasting. Sustain. Energy Technol. Assess. 2022, 52, 102275. [Google Scholar] [CrossRef]
- Amalou, I.; Mouhni, N.; Abdali, A. Multivariate time series prediction by RNN architectures for energy consumption forecasting. Energy Rep. 2022, 8, 1084–1091. [Google Scholar] [CrossRef]
- Lee, Y.; Ha, B.; Hwangbo, S. Generative model-based hybrid forecasting model for renewable electricity supply using long short-term memory networks: A case study of South Korea’s energy transition policy. Renew. Energy 2022, 200, 69–87. [Google Scholar] [CrossRef]
- Yang, B.; Yuan, X.; Tang, F. Improved nonlinear mapping network for wind power forecasting in renewable energy power system dispatch. Energy Rep. 2022, 8, 124–133. [Google Scholar] [CrossRef]
- Ren, H.; Li, Q.; Wu, Q.; Zhang, C.; Dou, Z.; Chen, J. Joint forecasting of multi-energy loads for a university based on copula theory and improved LSTM network. Energy Rep. 2022, 8, 605–612. [Google Scholar] [CrossRef]
- Ding, S.; Zhang, H.; Tao, Z.; Li, R. Integrating data decomposition and machine learning methods: An empirical proposition and analysis for renewable energy generation forecasting. Expert Syst. Appl. 2022, 204, 117635. [Google Scholar] [CrossRef]
- Khan, W.; Walker, S.; Zeiler, W. Improved solar photovoltaic energy generation forecast using deep learning-based ensemble stacking approach. Energy 2022, 240, 122812. [Google Scholar] [CrossRef]
- Chaturvedi, S.; Rajasekar, E.; Natarajan, S.; McCullen, N. A comparative assessment of SARIMA, LSTM RNN and Fb Prophet models to forecast total and peak monthly energy demand for India. Energy Policy 2022, 168, 113097. [Google Scholar] [CrossRef]
- Zheng, J.; Du, J.; Wang, B.; Klemeš, J.J.; Liao, Q.; Liang, Y. A hybrid framework for forecasting power generation of multiple renewable energy sources. Renew. Sustain. Energy Rev. 2023, 172, 113046. [Google Scholar] [CrossRef]
- Li, C.; Li, G.; Wang, K.; Han, B. A multi-energy load forecasting method based on parallel architecture CNN-GRU and transfer learning for data deficient integrated energy systems. Energy 2022, 259, 124967. [Google Scholar] [CrossRef]
- Rick, R.; Berton, L. Energy forecasting model based on CNN-LSTM-AE for many time series with unequal lengths. Eng. Appl. Artif. Intell. 2022, 113, 104998. [Google Scholar] [CrossRef]
- Shabbir, N.; Kütt, L.; Raja, H.A.; Jawad, M.; Allik, A.; Husev, O. Techno-economic analysis and energy forecasting study of domestic and commercial photovoltaic system installations in Estonia. Energy 2022, 253, 124156. [Google Scholar] [CrossRef]
- Khan, S.U.; Khan, N.; Ullah, F.U.M.; Kim, M.J.; Lee, M.Y.; Baik, S.W. Towards intelligent building energy management: AI-based framework for power consumption and generation forecasting. Energy Build. 2023, 279, 112705. [Google Scholar] [CrossRef]
- Niu, D.; Yu, M.; Sun, L.; Gao, T.; Wang, K. Short-term multi-energy load forecasting for integrated energy systems based on CNN-BiGRU optimized by attention mechanism. Appl. Energy 2022, 313, 118801. [Google Scholar] [CrossRef]
- Peng, L.; Wang, L.; Xia, D.; Gao, Q. Effective energy consumption forecasting using empirical wavelet transform and long short-term memory. Energy 2022, 238, 121756. [Google Scholar] [CrossRef]
- Kim, H.; Kim, M. A novel deep learning-based forecasting model optimized by heuristic algorithm for energy management of microgrid. Appl. Energy 2023, 332, 120525. [Google Scholar] [CrossRef]
- Khan, Z.A.; Ullah, A.; Ul Haq, I.; Hamdy, M.; Maria Mauro, G.; Muhammad, K.; Hijji, M.; Baik, S.W. Efficient Short-Term Electricity Load Forecasting for Effective Energy Management. Sustain. Energy Technol. Assessments 2022, 53, 102337. [Google Scholar] [CrossRef]
- Yan, K.; Zhou, X.; Chen, J. Collaborative deep learning framework on IoT data with bidirectional NLSTM neural networks for energy consumption forecasting. J. Parallel Distrib. Comput. 2022, 163, 248–255. [Google Scholar] [CrossRef]
- He, Y.L.; Chen, L.; Gao, Y.; Ma, J.H.; Xu, Y.; Zhu, Q.X. Novel double-layer bidirectional LSTM network with improved attention mechanism for predicting energy consumption. ISA Trans. 2022, 127, 350–360. [Google Scholar] [CrossRef] [PubMed]
- Hochreiter, S.; Schmidhuber, J. Long short-term memory. Neural Comput. 1997, 9, 1735–1780. [Google Scholar] [CrossRef]
- Wu, Y.; Schuster, M.; Chen, Z.; Le, Q.V.; Norouzi, M.; Macherey, W.; Krikun, M.; Cao, Y.; Gao, Q.; Macherey, K.; et al. Google’s neural machine translation system: Bridging the gap between human and machine translation. arXiv 2016, arXiv:1609.08144. [Google Scholar]
- Graves, A.; Schmidhuber, J. Offline handwriting recognition with multidimensional recurrent neural networks. In Proceedings of the Advances in Neural Information Processing Systems, San Francisco, CA, USA, 30 November–3 December 2008; Volume 21. [Google Scholar]
- Schmidhuber, J.; Wierstra, D.; Gomez, F.J. Evolino: Hybrid neuroevolution/optimal linear search for sequence prediction. In Proceedings of the 19th International Joint Conferenceon Artificial Intelligence (IJCAI), Scotland, UK, 30 July–5 August 2005. [Google Scholar]
- Schmidhuber, J. Deep learning in neural networks: An overview. Neural Netw. 2015, 61, 85–117. [Google Scholar]
- Siami-Namini, S.; Tavakoli, N.; Namin, A.S. The performance of LSTM and BiLSTM in forecasting time series. In Proceedings of the 2019 IEEE International Conference on Big Data (Big Data), Los Angeles, CA, 9–12 December 2019; pp. 3285–3292. [Google Scholar]
- Yang, X.S.; Deb, S.; Zhao, Y.X.; Fong, S.; He, X. Swarm intelligence: Past, present and future. Soft Comput. 2018, 22, 5923–5933. [Google Scholar] [CrossRef] [Green Version]
- Li, M.; Du, W.; Nian, F. An adaptive particle swarm optimization algorithm based on directed weighted complex network. Math. Probl. Eng. 2014, 2014, 434972. [Google Scholar] [CrossRef] [Green Version]
- Van, P.T.; Van, T.H.; Tangaramvong, S. Performance Comparison of Variants Based on Swarm Intelligence Algorithm of Mathematical and Structural Optimization. Iop Conf. Ser. Mater. Sci. Eng. 2022, 1222, 012013. [Google Scholar] [CrossRef]
- Cui, Y.; Jia, L.; Fan, W. Estimation of actual evapotranspiration and its components in an irrigated area by integrating the Shuttleworth-Wallace and surface temperature-vegetation index schemes using the particle swarm optimization algorithm. Agric. For. Meteorol. 2021, 307, 108488. [Google Scholar] [CrossRef]
- Priyadarshini, I.; Cotton, C. A novel LSTM–CNN–grid search-based deep neural network for sentiment analysis. J. Supercomput. 2021, 77, 13911–13932. [Google Scholar] [CrossRef] [PubMed]
- Thanh, T.; Van Dai, L.; Minh, L. Effects of Data Standardization on Hyperparameter Optimization with the Grid Search Algorithm Based on Deep Learning: A Case Study of Electric Load Forecasting. Adv. Technol. Innov. 2022, 7, 258–269. [Google Scholar] [CrossRef]
- Subramanian, S.; Rao, A. Deep-learning based Time Series Forecasting of Go-around Incidents in the National Airspace System. In Proceedings of the 2018 AIAA Modeling and Simulation Technologies Conference, Atlanta, GA, USA, 25–29 June 2018. [Google Scholar] [CrossRef]
- Stone, M. Cross-validatory choice and assessment of statistical predictions. J. R. Stat. Soc. Ser. B 1974, 36, 111–133. [Google Scholar] [CrossRef]
- Krizhevsky, A.; Sutskever, I.; Hinton, G.E. Imagenet classification with deep convolutional neural networks. Commun. ACM 2017, 60, 84–90. [Google Scholar] [CrossRef] [Green Version]
- Schölkopf, B.; Smola, A.J.; Bach, F. Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond; MIT Press: Cambridge, MA, USA, 2002. [Google Scholar]
- Suykens, J.A.; Vandewalle, J. Least squares support vector machine classifiers. Neural Process. Lett. 1999, 9, 293–300. [Google Scholar] [CrossRef]
- Liaw, A.; Wiener, M. Classification and regression by randomForest. R News 2002, 2, 18–22. [Google Scholar]
- Dorogush, A.V.; Ershov, V.; Gulin, A. CatBoost: Gradient boosting with categorical features support. arXiv 2018, arXiv:1810.11363. [Google Scholar]
- Ke, G.; Meng, Q.; Finley, T.; Wang, T.; Chen, W.; Ma, W.; Ye, Q.; Liu, T.Y. Lightgbm: A highly efficient gradient boosting decision tree. In Proceedings of the Advances in Neural Information Processing Systems, Long Beach, CA, USA, 4–9 December 2017; Volume 30. [Google Scholar]
- Chen, T.; Guestrin, C. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd Acm Sigkdd International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA; 2016; pp. 785–794. [Google Scholar]
Abbreviation | Definition | Expression |
---|---|---|
MAPE | Mean Absolute Percentage Error | |
MAE | Mean Absolute Error | |
RMSE | Root Mean Square Error | |
MSE | Mean Squared Error | |
MAAPE | Mean Arctangent Absolute Percentage Error | |
NRMSE | Normalized Root Mean Square Error | |
RMSPE | Root Mean Square Percentage Error | |
SMAPE | Symmetric Mean Absolute Percentage Error | |
U1 | Theil U Statistic 1 | |
U2 | Theil U Statistic 2 | |
IA | Index of Agreement | |
R | Coefficient of Determination |
CNN-BiLSTM | BiLSTM | CNN-LSTM | CNN | LSTM | SVR | LSSVR | RF | CatBoost | LGBM | XGBoost | |
---|---|---|---|---|---|---|---|---|---|---|---|
lr = 0.01 ℏ = 60 = 2 | lr = 0.1 ℏ = 20 | lr = 0.001 ℏ = 25 = 2 | lr = 0.1 = 2 | lr = 0.001 ℏ = 60 | C = 55 = = 0.03125 | C = 625 = 0.03125 | max_depth = 7 min_samples_leaf = 1 min_samples_split = 2 n_estimators = 50 | depth = 10 l2_leaf_reg = 100.0 lr = 0.01 | max_depth = 5 num_leaves = 50 reg_alpha = 0.01 reg_lambda = 1 | gamma = 0.0 lr = 0.1 max_depth = 5 min_child_weight = 1 reg_alpha = 1 | |
MAPE | 5.0417 | 5.8555 | 44.2395 | 22.0456 | 37.6811 | 20.8846 | 28.8258 | 5.1669 | 13.1710 | 5.9160 | 15.6317 |
MAE | 87.7030 | 102.3284 | 803.9523 | 428.3851 | 692.8386 | 343.4988 | 462.9180 | 91.8243 | 246.8932 | 105.8958 | 277.2572 |
RMSE | 110.8020 | 131.6346 | 869.9062 | 540.1201 | 767.0836 | 410.9314 | 571.3807 | 127.5901 | 318.3963 | 136.3436 | 371.4026 |
MSE | 12,277.0771 | 17,327.6641 | 756,736.8125 | 291,729.7188 | 588,417.2500 | 168,864.6442 | 326,475.9121 | 16,279.2250 | 101,376.2171 | 18,589.5845 | 137,939.8906 |
MAAPE | 0.0503 | 0.0584 | 0.4135 | 0.2132 | 0.3569 | 0.2011 | 0.2660 | 0.0514 | 0.1299 | 0.0590 | 0.1521 |
NRMSE | 0.0939 | 0.1115 | 0.7368 | 0.4575 | 0.6497 | 0.3481 | 0.4840 | 0.1081 | 0.2697 | 0.1155 | 0.3146 |
RMSPE | 0.0626 | 0.0727 | 0.4534 | 0.2593 | 0.3924 | 0.2577 | 0.3731 | 0.0708 | 0.1603 | 0.0730 | 0.2040 |
SMAPE | 5.0855 | 6.0200 | 57.8681 | 26.1360 | 47.5717 | 19.0293 | 24.1186 | 5.2382 | 13.8647 | 5.8904 | 16.0521 |
U1 | 0.0314 | 0.0375 | 0.3202 | 0.1746 | 0.2713 | 0.1123 | 0.1484 | 0.0360 | 0.0936 | 0.0383 | 0.1070 |
U2 | 0.0624 | 0.0741 | 0.4899 | 0.3042 | 0.4320 | 0.2314 | 0.3218 | 0.0719 | 0.1793 | 0.0768 | 0.2092 |
IA | 0.9720 | 0.9666 | 0.3690 | 0.4518 | 0.3935 | 0.3917 | 0.3762 | 0.9640 | 0.5246 | 0.9638 | 0.6462 |
0.8883 | 0.8423 | −5.8872 | −1.6551 | −4.3553 | −0.5369 | −1.9713 | 0.8518 | 0.0774 | 0.8308 | −0.2554 |
CNN-BiLSTM | BiLSTM | CNN-LSTM | CNN | LSTM | SVR | LSSVR | RF | CatBoost | LGBM | XGBoost | |
---|---|---|---|---|---|---|---|---|---|---|---|
lr = 0.01 ℏ = 50 = 2 | lr = 0.01 ℏ = 40 | lr = 0.01 ℏ = 65 = 3 | lr = 0.1 = 2 | lr = 0.1 ℏ = 45 | C = 5 = = 1.72844 | C = 15625 = 0.13446 | max_depth = 7 min_samples_leaf = 1 min_samples_split = 5 n_estimators = 50 | depth = 8 l2_leaf_reg = 0 lr = 0.01 | max_depth = 5 num_leaves = 50 reg_alpha = 0.001 reg_lambda = 0.1 | gamma = 0.0 lr = 0.5 max_depth = 5 min_child_weight = 1 reg_alpha = 1 | |
MAPE | 5.0292 | 5.3492 | 40.6992 | 21.2876 | 28.0344 | 11.0438 | 13.8404 | 9.0470 | 7.8271 | 7.3418 | 7.3984 |
MAE | 47.1754 | 50.2112 | 453.6898 | 227.9078 | 273.9267 | 111.5945 | 135.6584 | 89.3437 | 76.8737 | 72.0425 | 74.2770 |
RMSE | 68.3823 | 83.3169 | 555.3765 | 292.4198 | 339.5989 | 143.5384 | 166.9033 | 110.3985 | 97.0129 | 95.2964 | 96.7993 |
MSE | 4676.1406 | 6941.7031 | 308,443.0938 | 85,509.3438 | 115,327.4141 | 20,603.2779 | 27,856.7086 | 12,187.8240 | 9411.4937 | 9081.4098 | 9370.1064 |
MAAPE | 0.0501 | 0.0528 | 0.3781 | 0.2067 | 0.2621 | 0.1093 | 0.1366 | 0.0899 | 0.0779 | 0.0730 | 0.0736 |
NRMSE | 0.0676 | 0.0824 | 0.5491 | 0.2891 | 0.3358 | 0.1419 | 0.1650 | 0.1091 | 0.0959 | 0.0942 | 0.0957 |
RMSPE | 0.0692 | 0.0916 | 0.4404 | 0.2482 | 0.3513 | 0.1322 | 0.1622 | 0.1071 | 0.0926 | 0.0908 | 0.0918 |
SMAPE | 5.0129 | 5.6868 | 54.0280 | 23.7873 | 27.1095 | 11.8818 | 14.7027 | 9.5339 | 8.1516 | 7.5961 | 7.7344 |
U1 | 0.0329 | 0.0402 | 0.3546 | 0.1549 | 0.1689 | 0.0721 | 0.0848 | 0.0545 | 0.0476 | 0.0465 | 0.0476 |
U2 | 0.0660 | 0.0805 | 0.5364 | 0.2824 | 0.3280 | 0.1386 | 0.1612 | 0.1066 | 0.0937 | 0.0920 | 0.0935 |
IA | 0.9895 | 0.9840 | 0.4640 | 0.6817 | 0.5731 | 0.9491 | 0.9192 | 0.9717 | 0.9786 | 0.9795 | 0.9780 |
0.9546 | 0.9327 | −1.9923 | 0.1704 | −0.1188 | 0.8001 | 0.7298 | 0.8818 | 0.9087 | 0.9119 | 0.9091 |
CNN-BiLSTM | BiLSTM | CNN-LSTM | CNN | LSTM | SVR | LSSVR | RF | CatBoost | LGBM | XGBoost | |
---|---|---|---|---|---|---|---|---|---|---|---|
lr = 0.01 ℏ = 55 = 2 | lr = 0.01 ℏ = 25 | lr = 0.01 ℏ = 70 = 2 | lr = 0.01 = 2 | lr = 0.1 ℏ = 70 | C = 20 = = 0.04500 | C = 625 = 0.40171 | max_depth = 7 min_samples_leaf = 1 min_samples_split = 2 n_estimators = 100 | depth = 7 l2_leaf_reg = 1 lr = 0.01 | max_depth = 5 num_leaves = 50 reg_alpha = 0.01 reg_lambda = 1 | gamma = 0.0 lr = 0.5 max_depth = 3 min_child_weight = 2 reg_alpha = 0.01 | |
MAPE | 5.4774 | 12.6515 | 46.7455 | 27.8494 | 19.9738 | 15.4584 | 13.0736 | 13.0576 | 12.0870 | 14.0688 | 16.0561 |
MAE | 21.9951 | 47.7852 | 217.1153 | 93.8588 | 76.8539 | 69.5279 | 54.7455 | 56.1637 | 49.5175 | 56.9160 | 69.7948 |
RMSE | 27.1889 | 60.9706 | 273.3944 | 103.2772 | 90.8298 | 94.1978 | 74.6178 | 76.6894 | 65.7881 | 77.1039 | 96.7247 |
MSE | 739.2354 | 3717.4133 | 74,744.4766 | 10,666.1846 | 8250.0557 | 8873.2195 | 5567.8181 | 5881.2619 | 4328.0769 | 5945.0158 | 9355.6621 |
MAAPE | 0.0547 | 0.1248 | 0.4225 | 0.2650 | 0.1952 | 0.1515 | 0.1287 | 0.1285 | 0.1194 | 0.1382 | 0.1569 |
NRMSE | 0.0565 | 0.1267 | 0.5681 | 0.2146 | 0.1888 | 0.1958 | 0.1551 | 0.1594 | 0.1367 | 0.1602 | 0.2010 |
RMSPE | 0.0625 | 0.1545 | 0.5176 | 0.3233 | 0.2237 | 0.1887 | 0.1604 | 0.1611 | 0.1445 | 0.1719 | 0.1976 |
SMAPE | 5.5394 | 11.5774 | 66.3500 | 32.5211 | 20.7827 | 17.2753 | 14.5297 | 14.5490 | 13.2565 | 15.6626 | 18.3556 |
U1 | 0.0322 | 0.0680 | 0.4539 | 0.1319 | 0.1092 | 0.1204 | 0.0943 | 0.0975 | 0.0828 | 0.0972 | 0.1253 |
U2 | 0.0642 | 0.1439 | 0.6453 | 0.2438 | 0.2144 | 0.2223 | 0.1761 | 0.1810 | 0.1553 | 0.1820 | 0.2283 |
IA | 0.9930 | 0.9680 | 0.4525 | 0.8915 | 0.9243 | 0.8941 | 0.9408 | 0.9360 | 0.9545 | 0.9381 | 0.8952 |
0.9720 | 0.8590 | −1.8351 | 0.5954 | 0.6871 | 0.6634 | 0.7888 | 0.7769 | 0.8358 | 0.7745 | 0.6451 |
CNN-BiLSTM | BiLSTM | CNN-LSTM | CNN | LSTM | SVR | LSSVR | RF | CatBoost | LGBM | XGBoost | |
---|---|---|---|---|---|---|---|---|---|---|---|
lr = 0.01 ℏ = 125 = 2 | lr = 0.001 ℏ = 150 | lr = 0.01 ℏ = 5 = 2 | lr = 0.1 = 3 | lr = 0.01 ℏ = 10 | C = 100 = = 1.72844 | C = 390625 = 0.09336 | max_depth = 11 min_samples_leaf = 2 min_samples_split = 10 n_estimators = 100 | depth = 6 l2_leaf_reg = 1 lr = 0.01 | max_depth = 5 num_leaves = 50 reg_alpha = 1 reg_lambda = 1 | gamma = 0.0 lr = 0.05 max_depth = 3 min_child_weight = 1 reg_alpha = 1 | |
MAPE | 4.0034 | 18.0789 | 53.1444 | 24.0272 | 44.4674 | 9.0418 | 10.6892 | 11.9223 | 10.8338 | 12.0526 | 11.0120 |
MAE | 30.5129 | 138.1454 | 424.3060 | 176.0921 | 353.9534 | 75.3807 | 85.4468 | 95.7201 | 86.7298 | 96.8677 | 88.6650 |
RMSE | 37.4739 | 144.8520 | 449.4491 | 197.1012 | 374.8922 | 96.5099 | 111.9645 | 119.9051 | 106.6557 | 124.0479 | 106.3436 |
MSE | 1404.2955 | 20,982.1133 | 202,004.5000 | 38,848.8867 | 140,544.1250 | 9314.1634 | 12,536.0451 | 14,377.2259 | 11,375.4350 | 15,387.8716 | 11,308.9697 |
MAAPE | 0.0400 | 0.1784 | 0.4859 | 0.2318 | 0.4159 | 0.0898 | 0.1057 | 0.1178 | 0.1074 | 0.1190 | 0.1093 |
NRMSE | 0.0754 | 0.2913 | 0.9039 | 0.3964 | 0.7540 | 0.1941 | 0.2252 | 0.2412 | 0.2145 | 0.2495 | 0.2139 |
RMSPE | 0.0491 | 0.1885 | 0.5391 | 0.2751 | 0.4538 | 0.1097 | 0.1355 | 0.1446 | 0.1273 | 0.1472 | 0.1261 |
SMAPE | 4.0839 | 20.0636 | 73.4016 | 23.1385 | 58.0351 | 9.6521 | 11.0102 | 12.5102 | 11.4853 | 12.7886 | 11.5984 |
U1 | 0.0240 | 0.1005 | 0.3939 | 0.1240 | 0.3088 | 0.0641 | 0.0727 | 0.0788 | 0.0704 | 0.0818 | 0.0704 |
U2 | 0.0475 | 0.1836 | 0.5696 | 0.2498 | 0.4751 | 0.1223 | 0.1419 | 0.1520 | 0.1352 | 0.1572 | 0.1348 |
IA | 0.9840 | 0.7973 | 0.3550 | 0.4935 | 0.4142 | 0.8586 | 0.7996 | 0.8019 | 0.8464 | 0.7894 | 0.8249 |
0.9330 | −0.0004 | −8.6318 | −0.8523 | −5.7013 | 0.5559 | 0.4023 | 0.3145 | 0.4576 | 0.2663 | 0.4608 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Chen, Y.; Fu, Z. Multi-Step Ahead Forecasting of the Energy Consumed by the Residential and Commercial Sectors in the United States Based on a Hybrid CNN-BiLSTM Model. Sustainability 2023, 15, 1895. https://doi.org/10.3390/su15031895
Chen Y, Fu Z. Multi-Step Ahead Forecasting of the Energy Consumed by the Residential and Commercial Sectors in the United States Based on a Hybrid CNN-BiLSTM Model. Sustainability. 2023; 15(3):1895. https://doi.org/10.3390/su15031895
Chicago/Turabian StyleChen, Yifei, and Zhihan Fu. 2023. "Multi-Step Ahead Forecasting of the Energy Consumed by the Residential and Commercial Sectors in the United States Based on a Hybrid CNN-BiLSTM Model" Sustainability 15, no. 3: 1895. https://doi.org/10.3390/su15031895
APA StyleChen, Y., & Fu, Z. (2023). Multi-Step Ahead Forecasting of the Energy Consumed by the Residential and Commercial Sectors in the United States Based on a Hybrid CNN-BiLSTM Model. Sustainability, 15(3), 1895. https://doi.org/10.3390/su15031895