Wind power prediction based on outlier correction, ensemble reinforcement learning, and residual correction

Shi Yin and Hui Liu

Energy, 2022, vol. 250, issue C

Abstract: Wind power prediction contributes to clean energy utilization and grid dispatching. In this study, a wind power prediction model based on outlier correction, ensemble reinforcement learning, and residual correction is proposed. Firstly, the Hampel identifier (HI) is utilized to correct outliers in the original data. Then group method of data handling, echo state network, and extreme learning machine are selected as three base learners to predict the corrected wind power data. And the ensemble reinforcement learning algorithm based on the Q-learning algorithm is utilized to generate optimal ensemble weights to combine the prediction results of three base learners. Finally, the residual correction method (RCM) is applied to revise the prediction results, to obtain the final forecasting results. By comparing the experimental results of the relevant models for four real wind power datasets, it can be known that: (a) The use of both HI and RCM is beneficial to enhance the model's prediction accuracy. (b) The proposed ensemble method based on the Q-learning algorithm has superiority and can achieve smaller prediction errors than three classic ensemble algorithms. (c) The wind power prediction model proposed in this paper has excellent prediction performance and is superior to five commonly used intelligent models.

Keywords: Wind power prediction; Outlier correction; Ensemble reinforcement learning; Residual correction method (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (10)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544222007605
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:250:y:2022:i:c:s0360544222007605

DOI: 10.1016/j.energy.2022.123857

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 ().