Abstract
Time-series extrapolation which is also known as univariate time series forecasting relies on quantitative methods to analyse data for the variable of interest. Pure extrapolation is based only on values of variable being forecast. We are interested in forecasting the electricity generated for Malaysia. The Tenaga Nasional Berhad (TNB) operates an electricity network with the largest capacity of over 7100MW that accounts for over 62\% of the total power generation of Peninsular Malaysia. The rest of the power is generated by other Independent Power Producer (IPP). A forecasting model has been developed which identifies seasonal factors in the time-series. Seasonality often accounts for the major part of time series data. In this paper we examine the forecasting performance of Box-Jenkins methodology for SARIMA models and ARIMA models to forecast future electricity generated for Malaysia. We employ the data on the electricity generated at Power Plant to forecast future electricity demand. The error statistics of forecast between the models for a month ahead are presented and the behaviour of data is also observed. Keywords: Forecasting; electricity demand; SARIMA, Box Jenkins; genetic algorithm.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have