Modeling Wind Speed Data Using Dynamic Linear Model and Innovation State Space ETS Model
Keywords:
Wind speed, Dynamic Linear Model, ETS state space model, Kalman filter Ljung Box testAbstract
Study’s Excerpt/Novelty
- This research introduces an application of dynamic linear models (DLM) and Innovation state space models based on Exponential smoothing methods to effectively capture and forecast wind speed patterns in Katsina metropolis from January 2013 to December 2022.
- By incorporating trend, seasonal, and regressive components, the study provides a comprehensive and interpretable modeling framework.
- The DLM's superior performance in one-step-ahead forecasts, validated through residual analysis, demonstrates its robustness and suitability for accurately replicating the observed wind speed data, offering valuable insights for wind speed prediction in complex systems.
Full Abstract
For many years, experts have acknowledged the intricate nature of wind speed, considering it a prime example of complex systems. In recent times, wind speed patterns have become increasingly erratic. This research paper focuses on a state space approach that utilizes dynamic linear models (DLM) and an Innovation state space model based on Exponential smoothing methods to effectively model wind speed data in the Katsina metropolis from January 2013 to December 2022. By incorporating trend, seasonal, and regressive components, these methods allow for a natural interpretation of the data. The model's parameters were estimated, and their validity was assessed through residual analysis. The validated models were then used for one-step-ahead forecasts. Comparing the predicted values with the observed wind speed series, the DLM demonstrates a closer match than the Innovation state space ETS (M, N, A), indicating that the model accurately replicates the actual data. Consequently, the model is considered suitable for representing the wind speed of Katsina.
References
Agada I.O., Peter O. and Agada P.O, (2023). Stochastic Modelling of Wind Speed over Northern States in Nigeria Science World Journal Vol. 18(No 3) DOI: https://dx.doi.org/10.4314/swj.v18i3.11
Azorin-Molina, C., S. Rehman, J. A. Guijarro, T. R. McVicar, L. Minola, D. Chen, and S. M. Vicente-Serrano, (2018) Recent trends in wind speed across Saudi Arabia, 1978–2013: A break in the stilling. Int. J. Climatol., 38, e966–e984, https://doi.org/10.1002/joc.5423.
Bessa, R. J., Miranda, V., Botterud, A., Wang, J., & Zhang, J. (2020). Wind power forecasting: A review of the state of the art. IEEE Transactions on Power Systems, 35(1), 425-440.
Chai, Tianyou, et al. (2019)."A review on hybrid forecasting approaches: Principles, applications, and challenges." Renewable and Sustainable Energy Reviews 113: 109300.
Chatterjee, S., & Mondal, S. (2021). Wind speed and power forecasting using hybrid machine learning models: a comprehensive review. Renewable Energy, 176, 624-641.
Chen, X., Wang, Y., Liu, P., & Liu, S. (2019). A Dynamic Linear Model for Monthly Rainfall Prediction Based on the Bayesian Inference Method. Journal of Hydrology, 577, 123998.
Dare, J., Patrick, A. O., & Oyewola, D. O. (2022). Comparison of Stationarity on Ljung Box Test Statistics for Forecasting. Earthline Journal of Mathematical Sciences, 8(2), 325-336. https://doi.org/10.34198/ejms.8222.325336
Hyndman, R.J. and Khandakar, Y. (2008). Automatic Time Series Forecasting: The forecast Package for R. Journal of statistical software. https://doi.org/10.18637/jss.v027.i03.
Hyndman, R.J., Koehler, A.B., Ord, J.K., Snyder, R.D. (2008). Forecasting with exponential smoothing: the state space approach. Springer-Verlag, Berlin. https://doi.org/10.1007/978-3-540-71918-2
Jiang, Y., Y. Luo, and Z. Zhao (2013). Maximum wind speed changes over China. Acta Meteor. Sin., 27, 63–74, https://doi.org/10.1007/s13351-013-0107-x.
Marangon L.M. Lima, E. Popova, P. Damien (2014). Modeling and forecasting of Brazilian reservoir inflows via dynamic linear models, International Journal of Forecasting, Volume 30, Issue3, Pages 464-476, ISSN 0169-2070, https://doi.org/10.1016/j.ijforecast.2013.12.009.
Nobre, F. F., Monteiro, A. B. S., Telles, P. R., & Williamson, G. D. (2001). Dynamic linear model and SARIMA: a comparison of their forecasting performance in epidemiology. Statistics in medicine, 20(20), 3051-3069. https://doi.org/10.1002/sim.963
Osthus, D., Caragea, P. C., Higdon, D., Morley, S. K., Reeves, G. D., & Weaver, B. P. (2014). Dynamic linear models for forecasting of radiation belt electrons and limitations on physical interpretation of predictive models. Space Weather, 12(6), 426–446. http://doi.org/10.1002/2014SW001057
Petris, G. (2010). An R Package for Dynamic Linear Models. Journal of Statistical Software, 36(12), 1–16. http://doi.org/10.1007/BF02887432
Wang, Xuan, et al. (2019). "A hybrid model for wind speed forecasting based on LSTM and ARIMA." Applied Energy 233-234: 124-135.
Wang, Y., Huang, S., Zhou, L., Wang, H., & Li, Z. (2022). A deep learning approach for wind speed forecasting based on long short-term memory neural networks. Energy Reports, 8, 333-342.
ZhangG, C. Azorin-Molina, D. Chen, J.A. Guijarro, F. Kong, L. Minola, T.R. McVicar, S.-W. Son, P. Shi (2020). Variability of daily maximum wind speed across China, 1975–2016: an examination of likely causes J. Clim., 33, pp. 2793-2816. https://doi.org/10.1175/JCLI-D-19-0603.1
Zhang, J., Cheng, W., Tang, M., Jiang, X., & Wang, G. (2021). A review of wind speed forecasting methods based on machine learning. Renewable and Sustainable Energy Reviews, 135, 110159
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 UMYU Scientifica
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
&
