A wind power ultra-short-term ensemble forecast framework considering wind speed correction and scenario classification

Congcong Chen, Lin Chai and Qingling Wang

Energy, 2025, vol. 336, issue C

Abstract: The volatility, intermittency and uncertainty of wind power make it difficult to match it with electricity demand in real time. Accurate wind power prediction (WPP) is the basis for improving wind energy consumption. This paper proposes a wind power ultra-short-term ensemble forecast method based on wind speed correction and forecast scenario classification. Firstly, the given prediction wind speed correction algorithm (GPWSCA) is designed. The wind speed deviation (WSD) is decomposed by ICEEDMAN and reconstructed into fluctuation and trend quantities. The trend quantities are predicted by RBF, and the BiLSTM-FPS-DA-LSTM prediction model is designed to predict the fluctuation. Secondly, the WPP scenario partition algorithm (WPPSPA) is proposed, which includes initial scenario partition for fluctuation range and secondary partition for fluctuation trend. On this basis, in order to solve the problem of limited data in some scenarios after scenario partition, an ultra-short-term stacking ensemble forecast model based on TimeGAN data augmentation technology (TimeGAN-USEFM) is designed to enhance WPP accuracy by increasing data diversity and structural diversity. Finally, multiple sets of comparative experiments are conducted using real data from a wind farm in Southwest China to evaluate the proposed GPWSCA, WPPSPA, and TimeGAN-USEFM. The results indicate that using WPPSPA improves average R2 prediction performance by 0.167, while MAE and RMSE decrease by 0.434 and 0.398, respectively. TimeGAN-USEFM shows an average R2 improvement of 0.063 over USEFM, with MAE and RMSE decreasing by 0.157 and 0.154, respectively, validating the feasibility and effectiveness of the proposed GPWSCA, WPPSPA, and TimeGAN-USEFM in the field of WPP.

Keywords: Wind power prediction; Ultra-short-term forecasting; Given prediction wind speed correction; Scenario partitioning; Time series generative adversarial network; Stacking ensemble forecasting (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544225039180
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:336:y:2025:i:c:s0360544225039180

DOI: 10.1016/j.energy.2025.138276

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().