FORECASTING DAILY ELECTRICITY DEMAND FOR TURKEY

Authors : Ebru YÜKSEL HALİLOĞLU, Behzat Ecem TUTU

Pages : 40-49

View : 23 | Download : 8

Publication Date : 2018-12-12

Article Type : Research Paper

Abstract :The aim of this study is to enhance a time series model that estimates daily electricity energy demand for Turkey. In literature, annual and monthly electricity demand has been forecasted to take long term decision mostly, this study distinctively makes daily predictions to help daily operations for Turkish electricity market. Using Box-Jenkins methodology, an ARIMA model is constructed to forecast daily electricity demand for Turkey.The best model is found to be SARIMAinsert ignore into journalissuearticles values(1,1,0);insert ignore into journalissuearticles values(1,0,1);7 to estimate the consumption for a day later. According to this time series model, Turkey daily electricity demand in January and February of 2018 is forecasted about 1.45% deviation. As the data is limited, only the first two month of 2018 can be compared to the actual daily consumption. This study will be carried on for the purpose of validating the model for other months of 2018.
Keywords : ARIMA, Electricity demand, Forecast