Title :
Wind power ultra-short-term forecasting method combined with pattern-matching and ARMA-model
Author :
Yunlong Cao ; Yanhua Liu ; Dongying Zhang ; Wei Wang ; Zhenhuan Chen
Author_Institution :
Sch. of Electr. Eng., North China Electr. Power Univ., Beijing, China
Abstract :
In accordance with 0~6 hour ultra-short-term forecasting of wind power, a method combine two ways of forecasting is put forward in this paper, in this method, Pattern-matching is used to make a 0~24 hour forecasting by taking the data of the day before the forecasting day as input, ARMA-model forecasting is used to make a 0~1 hour forecasting, combine the two results of forecasting to get a 0~6 hour forecasting result. The example shows this method can improve the accuracy of ultra-short-term wind power forecasting compared to traditional ARMA-model forecasting.
Keywords :
autoregressive moving average processes; load forecasting; wind power plants; ARMA-model forecasting; pattern-matching; time 0 hour to 24 hour; wind power ultrashort-term forecasting method; Accuracy; Data models; Forecasting; History; Predictive models; Wind forecasting; Wind power generation; ARMA-model; Pattern-matching; forecasting; wind power;
Conference_Titel :
PowerTech (POWERTECH), 2013 IEEE Grenoble
Conference_Location :
Grenoble
DOI :
10.1109/PTC.2013.6652257