Wind speed interval prediction model based on variational mode decomposition and multi-objective optimization

As a potential new energy power generation technology, wind power is gradually developing into the world’s mainstream energy. In the research on wind power generation, wind speed prediction is an important part, which has been widely studied. The accurate wind speed prediction is a key part of wind...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Applied soft computing Ročník 113; s. 107848
Hlavní autoři: Wang, Jianzhou, Cheng, Zishu
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.12.2021
Témata:
Combined models
Interval prediction
Multi-objective optimization algorithm
Wind speed
Multi-objective optimization algorithm
Combined models
Wind speed
Interval prediction
ISSN:1568-4946, 1872-9681
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:As a potential new energy power generation technology, wind power is gradually developing into the world’s mainstream energy. In the research on wind power generation, wind speed prediction is an important part, which has been widely studied. The accurate wind speed prediction is a key part of wind power management to help wind power grid-tied. Currently, most research has focused on point prediction, which in fact does not facilitate the quantitative characterization of the endogenous uncertainty involved. However, interval prediction can avoid this deficiency and make better operation and scheduling of the wind power models. In this study, a novel interval prediction model based on wind speed distribution and multi-objective optimization is designed, which includes data noise reduction module, prediction module, and combination module based on a multi-objective salp swarm algorithm, to provide accurate forecast for power model operation and grid dispatching. The 10-minute wind speed data from three data sets in China were selected for prediction to evaluate the effectiveness of the proposed combined model. The results show that the model is not only better than the considered benchmark model, but also has good potential practical application value in wind power models. •Developed the novel combined model on data ensemble decomposition and reconstruction and a swarm intelligence-based algorithm.•Provide an effective method for choosing the de-noising model.•Interval prediction of wind speed can better help power grid connection and dispatch.
ISSN:1568-4946
1872-9681
DOI:10.1016/j.asoc.2021.107848