A Hybrid Model of Decomposition, Extended Kalman Filter and Autoregressive - Long Short - Term Memory Network for Hourly Day Ahead Wind Speed Forecasting

Nguyen Thi Hoai  Thu; Phan Quoc  Bao; Pham Nang  Van

doi:10.6180/jase.202409_27(9).0004

A Hybrid Model of Decomposition, Extended Kalman Filter and Autoregressive - Long Short - Term Memory Network for Hourly Day Ahead Wind Speed Forecasting

Nguyen Thi Hoai Thu¹This email address is being protected from spambots. You need JavaScript enabled to view it., Phan Quoc Bao², and Pham Nang Van¹

¹Power Grid and Renewable Energy Lab., School of Electrical and Electronic Engineering, Hanoi University of Science and Technology, Vietnam

²School of Informatics, Computing and Cyber security, Northern Arizona University, USA

Received: June 30, 2023
Accepted: November 17, 2023
Publication Date: December 6, 2023

Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Download Citation: ||https://doi.org/10.6180/jase.202409_27(9).0004

This research work introduced an innovative forecasting approach that combined the Autoregressive – Long Short-Term Memory (AR-LSTM) neural network with decomposition techniques and the Extended Kalman Filter (EKF) to predict hourly day-ahead wind speed. The process started with data pre-processing, followed by decomposition into three distinct components: trend, seasonal, and residual, using the Seasonal and Trend decomposition using Loess (STL) filter. The forecasting process was designed to handle each of these decomposed components independently. The trend and seasonal components were forecasted using the AR model, utilizing historical patterns and temporal dependencies. On the other hand, the residuals were predicted by a Long Short-Term Memory network, optimized through the application of the Extended Kalman Filter to improve the filtering process. Predictions from these individual components were then combined to generate the final wind speed forecast. To validate the proposed method, it was applied to real-world wind speed datasets from both Hanoi and Tokyo. The model’s performance was systematically compared with alternative methods. The results consistently demonstrated the superiority of the proposed approach over the three alternative methods, as evidenced by the Mean Absolute Error (MAE), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE) metrics. Impressively low values of MAE and MSE, along with an impressive MAPE value, were achieved, namely 6.79% and 10.3%, for hourly day-ahead wind speed prediction in Hanoi and Tokyo, respectively. These findings underscore the robustness and effectiveness of the proposed model in delivering highly accurate wind speed predictions for both geographical locations.

Keywords: Wind speed forecast; Hybrid model; Decomposition; Autoregressive – Long short-term memory; Extended Kalman Filter

[1] Y. Wang, R. Zou, F. Liu, L. Zhang, and Q. Liu, (2021) “A Review of Wind Speed and Wind Power Forecasting with Deep Neural Networks" Applied Energy 304: 117766. DOI: 10.1016/j.apenergy.2021.117766.
[2] N. T. H. Thu, P. N. Van, N. V. N. Nam, and P. H. Minh, (2022) “Forecasting Wind Speed Using A Hybrid Model Of Convolutional Neural Network And Long-Short Term Memory With Boruta Algorithm-Based Feature Selection" Journal of Applied Science and Engineering 26(8): 1053–1060. DOI: 10.6180/jase.202308_26(8).0001.
[3] A. Gupta, A. Kumar, and K. Boopathi, (2021) “Intraday Wind Power Forecasting Employing Feedback Mechanism" Electric Power Systems Research 201: 107518. DOI: 10.1016/j.epsr.2021.107518.
[4] Z. Zhen, G. Qiu, S. Mei, F. Wang, X. Zhang, R. Yin, Y. Li, G. J. Osório, M. Shafie-khah, and J. P. Catalão, (2022) “An Ultra-Short-Term Wind Speed Forecasting Model Based on Time Scale Recognition and Dynamic Adaptive Modeling" International Journal of Electrical Power & Energy Systems 135: 107502. DOI: 10.1016/j.ijepes.2021.107502.
[5] N. T. H. Thu, P. N. Van, and P. Q. Bao. “MultiStep Ahead Wind Speed Forecasting Based on a Bi-LSTM Network Combined with Decomposition Technique”. In: Computational Intelligence Methods for Green Technology and Sustainable Development. Ed. by Y.-P. Huang, W.-J. Wang, H. A. Quoc, H.-G. Le, and H.-N. Quach. Cham: Springer International Publishing, 2023, 569–580.
[6] Y.-K. Wu, J.-J. Zeng, G.-L. Lu, S.-W. Chau, and Y.-C. Chiang, (2020) “Development of an Equivalent Wind Farm Model for Frequency Regulation" IEEE Transactions on Industry Applications 56(3): 2360–2374. DOI: 10.1109/TIA.2020.2974418.
[7] Y. Wang, Q. Hu, L. Li, A. M. Foley, and D. Srinivasan, (2019) “Approaches to Wind Power Curve Modeling: A Review and Discussion" Renewable and Sustainable Energy Reviews 116: 109422. DOI: 10.1016/j.rser.2019.109422.
[8] T. H. T. Nguyen and Q. B. Phan, (2022) “Hourly Day Ahead Wind Speed Forecasting Based on a Hybrid Model of EEMD, CNN-Bi-LSTM Embedded with GA Optimization" Energy Reports 8: 53–60. DOI: 10.1016/j.egyr.2022.05.110.
[9] J. Yan and T. Ouyang, (2019) “Advanced Wind Power Prediction Based on Data-Driven Error Correction" Energy Conversion and Management 180: 302–311. DOI: 10.1016/j.enconman.2018.10.108.
[10] J. Jung and R. P. Broadwater, (2014) “Current Status and Future Advances for Wind Speed and Power Forecasting" Renewable and Sustainable Energy Reviews 31: 762–777. DOI: 10.1016/j.rser.2013.12.054.
[11] N. E. Huang and Z. Wu, (2008) “A Review on HilbertHuang Transform: Method and Its Applications to Geophysical Studies" Reviews of Geophysics 46(2): DOI: 10.1029/2007RG000228.
[12] J. Torres, A. García, M. De Blas, and A. De Francisco, (2005) “Forecast of Hourly Average Wind Speed with ARMA Models in Navarre (Spain)" Solar Energy 79(1): 65–77. DOI: 10.1016/j.solener.2004.09.013.
[13] K. Yunus, T. Thiringer, and P. Chen, (2016) “ARIMABased Frequency-Decomposed Modeling of Wind Speed Time Series" IEEE Transactions on Power Systems 31(4): 2546–2556. DOI: 10.1109/TPWRS.2015.2468586.
[14] R. C. Deo, M. A. Ghorbani, S. Samadianfard, T. Maraseni, M. Bilgili, and M. Biazar, (2018) “MultiLayer Perceptron Hybrid Model Integrated with the Firefly Optimizer Algorithm for Windspeed Prediction of Target Site Using a Limited Set of Neighboring Reference Station Data" Renewable Energy 116: 309–323. DOI: 10.1016/j.renene.2017.09.078.
[15] Y. Zhang, C. Zhang, Y. Zhao, and S. Gao, (2018) “Wind speed prediction with RBF neural network based on PCA and ICA" Journal of Electrical Engineering 69(2): 148–155. DOI: 10.2478/jee-2018-0018.
[16] E. Cadenas and W. Rivera, (2009) “Short Term Wind Speed Forecasting in La Venta, Oaxaca, México, Using Artificial Neural Networks" Renewable Energy 34(1): 274–278. DOI: 10.1016/j.renene.2008.03.014.
[17] J. Song, J. Wang, and H. Lu, (2018) “A Novel Combined Model Based on Advanced Optimization Algorithm for Short-Term Wind Speed Forecasting" Applied Energy 215: 643–658. DOI: 10.1016/j.apenergy.2018.02.070.
[18] J. Shi, J. Guo, and S. Zheng, (2012) “Evaluation of Hybrid Forecasting Approaches for Wind Speed and Power Generation Time Series" Renewable and Sustainable Energy Reviews 16(5): 3471–3480. DOI: 10.1016/j.rser.2012.02.044.
[19] C. Wang, H. Zhang, and P. Ma, (2020) “Wind Power Forecasting Based on Singular Spectrum Analysis and a New Hybrid Laguerre Neural Network" Applied Energy 259: 114139. DOI: 10.1016/j.apenergy.2019.114139.
[20] K. Zhang, R. Gençay, and M. Ege Yazgan, (2017) “Application of Wavelet Decomposition in Time-Series Forecasting" Economics Letters 158: 41–46. DOI: 10.1016/j.econlet.2017.06.010.
[21] Y. Zhao, L. Ye, W. Wang, H. Sun, Y. Ju, and Y. Tang, (2018) “Data-Driven Correction Approach to Refine Power Curve of Wind Farm Under Wind Curtailment" IEEE Transactions on Sustainable Energy 9(1): 95–105. DOI: 10.1109/TSTE.2017.2717021.
[22] X. Hou and L. Zhang. “Saliency Detection: A Spectral Residual Approach”. In: 2007 IEEE Conference on Computer Vision and Pattern Recognition. 2007 IEEE Conference on Computer Vision and Pattern Recognition. 2007, 1–8. DOI: 10.1109/CVPR.2007.383267.
[23] H. Ren, B. Xu, Y. Wang, C. Yi, C. Huang, X. Kou, T. Xing, M. Yang, J. Tong, and Q. Zhang. “Time-Series Anomaly Detection Service at Microsoft”. In: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. Comment: KDD 2019. 25, 2019, 3009–3017. DOI: 10.1145/3292500.3330680. arXiv: 1906.03821 [cs, stat].
[24] 6.6 STL Decomposition | Forecasting: Principles and Practice (2nd Ed). Accessed: Jun. 22, 2022.
[25] O. Trull, J. C. García-Díaz, and A. Peiró-Signes, (2022) “Multiple Seasonal STL Decomposition with DiscreteInterval Moving Seasonalities" Applied Mathematics and Computation 433: 127398. DOI: 10.1016/j.amc.2022.127398.
[26] C.-N. Ko and C.-M. Lee, (2013) “Short-Term Load Forecasting Using SVR (Support Vector Regression)-Based Radial Basis Function Neural Network with Dual Extended Kalman Filter" Energy 49: 413–422. DOI: 10.1016/j.energy.2012.11.015.
[27] N. T. Hoai Thu, P. N. Van, P. Q. Bao, N. V. Nhat Nam, P. H. Minh, and T. N. Quang. “Short-Term Forecasting of Solar Radiation Using a Hybrid Model of CNN-LSTM Integrated with EEMD”. In: 2022 6th International Conference on Green Technology and Sustainable Development (GTSD). 2022 6th International Conference on Green Technology and Sustainable Development (GTSD). 2022, 854–859. DOI: 10.1109/GTSD54989.2022.9988761.
[28] J. Qu, Z. Qian, and Y. Pei, (2021) “Day-Ahead Hourly Photovoltaic Power Forecasting Using Attention-Based CNN-LSTM Neural Network Embedded with Multiple Relevant and Target Variables Prediction Pattern" Energy 232: 120996. DOI: 10.1016/j.energy.2021.120996.
[29] N. N. V. Nhat, D. N. Huu, and T. N. T. Hoai, (2023) “Evaluating the EEMD-LSTM Model for Short-Term Forecasting of Industrial Power Load: A Case Study in Vietnam" International Journal of Renewable Energy Development 12(5): 881–890. DOI: 10.14710/ijred.2023.55078.
[30] T. H. T. Nguyen, Q. B. Phan, V. N. N. Nguyen, and H. M. Pham. “Day-Ahead Electricity Load Forecasting Based on Hybrid Model of EEMD and Bidirectional LSTM”. In: The 5th International Conference on Future Networks & Distributed Systems. ICFNDS 2021. New York, NY, USA: Association for Computing Machinery, 13, 2022, 31–41. DOI: 10.1145/3508072.3508079.