EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Short-Term Wind Power Forecasting Using an HOA–VMD–CNN–iTransformer–LSTM Deep Learning Framework

Huanjun Chen, Yan Wang, Jingran Sheng, Xinfu Pang iD, Zedong Zheng iD

Sensors · 06.11.2026 · 26 (21) · 7028

DOI: 10.3390/s26217028 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 06.11.2026. Verfasst von Huanjun Chen und 4 weiteren Autor:innen. Beteiligte Einrichtung: Key Laboratory of Energy Saving and Controlling in Power System of Liaoning Province, Shenyang Institute of Engineering, Shenyang 110136, China. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Department of Education of Liaoning Province of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable short-term wind power forecasts are difficult to obtain because power records are strongly non-stationary and hybrid networks are highly sensitive to parameter choices. This study proposes a Pearson–HOA–VMD–CNN–iTransformer–LSTM framework to address these challenges. Pearson coefficients are used to screen the candidate meteorological inputs. The hiking optimization algorithm (HOA) subsequently searches the mode number and penalty parameter of variational mode decomposition (VMD), together with the principal hyperparameters of the forecasting network. The optimized decomposition reduces sequence complexity, while a convolutional neural network (CNN) extracts local fluctuation patterns. iTransformer captures global variable dependencies using variate-wise self-attention, and long short-term memory (LSTM) models learn long-term temporal information. The framework was evaluated using four datasets containing 2976, 2880, 2976, and 2976 records, respectively, drawn from 2020 operational data from two Chinese wind farms with installed capacities of 99 and 200 MW. The sampling interval was 15 min, and the forecast horizons were 15, 30, and 45 min. Tests on four datasets show consistent gains over CNN–iTransformer–LSTM. Averaged across the datasets at the 15-min forecast horizon, RMSE, MAE, and MAPE decrease by 67.48%, 71.17%, and 49.66%, respectively, and mean R2 rises from 0.9672 to 0.9974.

Schlagwörter

Stichwörter der Autor:innen

LSTMhyperparameter optimizationiTransformershort-term wind power forecastingvariational mode decomposition

Autor:innen und Institutionen

  • Key Laboratory of Energy Saving and Controlling in Power System of Liaoning Province, Shenyang Institute of Engineering, Shenyang 110136, China.
  • School of Computing and Mathematical Sciences, University of Leicester, Leicester LE1 7RH, UK.

ORCID-Profile: Xinfu Pang, Zedong Zheng

Förderung

  • Department of Education of Liaoning Province of China

Zitieren

APA

Chen, H., Wang, Y., Sheng, J., Pang, X., & Zheng, Z. (2026). Short-Term Wind Power Forecasting Using an HOA–VMD–CNN–iTransformer–LSTM Deep Learning Framework. Sensors, 26(21), 7028. https://doi.org/10.3390/s26217028

Vancouver

Chen H, Wang Y, Sheng J, Pang X, Zheng Z. Short-Term Wind Power Forecasting Using an HOA–VMD–CNN–iTransformer–LSTM Deep Learning Framework. Sensors. 2026;26(21):7028. doi:10.3390/s26217028

BibTeX
@article{Chen2026,
  author = {Chen, Huanjun and Wang, Yan and Sheng, Jingran and Pang, Xinfu and Zheng, Zedong},
  title = {Short-Term Wind Power Forecasting Using an HOA–VMD–CNN–iTransformer–LSTM Deep Learning Framework},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {7028},
  doi = {10.3390/s26217028}
}
Kostenlos starten →