{"title":"Wind power Prediction based on the fusion of CN N-GRU combined neural network and attention mechanism","authors":"Yankun Wang, Zhenda Song, Yuchen Liu, Jiacheng Chen","doi":"10.1109/IAEAC54830.2022.9929869","DOIUrl":null,"url":null,"abstract":"Accurate prediction of wind power plays an i mportant role in long-term stable and safe operation of pow er system. Wind power generation can save a lot of energy a nd solve the problem of high cost and low utilization rate of traditional thermal power generation. In this paper, a comb ination of CNN-GRU-Attention model is proposed, and thre e different data decomposition methods are used for data pr e-processing, and the optimal data pre-processing method is selected for the model. Meanwhile, the fusion model is comp ared with CNN-LSTM-ATTENTION, GRU, RNN and other traditional single models. By analyzing MAE, RSME, R an d other evaluation indexes, it is concluded that the predictio n accuracy of CNN-GRU-Attention model is the best. In wi nd power prediction, the predicted value of the model is clos est to the actual value, and the prediction effect is higher tha n LSTM, BPNN and other traditional networks, which is of great significance for wind turbine power prediction.","PeriodicalId":349113,"journal":{"name":"2022 IEEE 6th Advanced Information Technology, Electronic and Automation Control Conference (IAEAC )","volume":"118 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE 6th Advanced Information Technology, Electronic and Automation Control Conference (IAEAC )","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IAEAC54830.2022.9929869","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Accurate prediction of wind power plays an i mportant role in long-term stable and safe operation of pow er system. Wind power generation can save a lot of energy a nd solve the problem of high cost and low utilization rate of traditional thermal power generation. In this paper, a comb ination of CNN-GRU-Attention model is proposed, and thre e different data decomposition methods are used for data pr e-processing, and the optimal data pre-processing method is selected for the model. Meanwhile, the fusion model is comp ared with CNN-LSTM-ATTENTION, GRU, RNN and other traditional single models. By analyzing MAE, RSME, R an d other evaluation indexes, it is concluded that the predictio n accuracy of CNN-GRU-Attention model is the best. In wi nd power prediction, the predicted value of the model is clos est to the actual value, and the prediction effect is higher tha n LSTM, BPNN and other traditional networks, which is of great significance for wind turbine power prediction.