IDHNet: Ultra‐Short‐Term Wind Power Forecasting With IVMD–DCInformer–HSSA Network

Autor: Wei Li, Lu Gao, Fei Zhang, XiaoYing Ren, Ling Qin
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: Energy Science & Engineering, Vol 12, Iss 12, Pp 5566-5589 (2024)
Druh dokumentu: article
ISSN: 2050-0505
24706248
DOI: 10.1002/ese3.1968
Popis: ABSTRACT The variability and unpredictability of wind power generation present significant challenges for grid management and planning. Enhancing the accuracy of wind power forecasting is crucial for improving the reliability of renewable energy systems. To enhance the accuracy of temporal wind power predictions, the IVMD–DCInformer–HSSA framework has been introduced. Initially, the original wind power data is decomposed into multiple intrinsic mode function (IMF) components using the improved variational mode decomposition (IVMD) technique. Subsequently, the Sparrow search algorithm (HSSA) is employed to optimize the parameters of the enhanced Informer deep neural network, which are then integrated into the improved Informer model. The predictions of each IMF component resulting from the IVMD decomposition are then combined to generate the final prediction outcome. The experimental results show that the R‐squared value of the proposed combined model is increased to 0.9903, and the accuracy is increased by 1%–3% compared with other models, which has a good prediction effect.
Databáze: Directory of Open Access Journals
Externí odkaz: