117 Dynamic Regression Model for Analyzing the Holidays' Effects to the Malaysian Load
Download citation file:
- Ris (Zotero)
- Reference Manager
The errors for the holidays in the short term load forecasting are known to be higher than those for weekends. Many studies have been done to improve the forecasting errors using various methods. This paper aims to find out which model would be a better model to estimate the moving holiday effects and therefore give a better forecasting accuracy for the peak weekly and daily load in Malaysia. Moving holiday effects in Malaysia are from the major festivals namely Aidil Fitri, Chinese New Year and Deepavali. Dynamic Regression (DR) or Linear Transfer function (LTF) modeling is considered and the final selection of the models depends on the mean absolute percentage error (MAPE) and others such as the sample autocorrelation function (ACF), the sample partial autocorrelation Function (PACF) and a Ljung-Box Chi-squared statistic test. The most appropriate non seasonal DR model for the weekly load to be used for forecasting would be the LTF (1, 1, 0) for the moving holidays and ARIMA(0, 1, 1) for the stochastic disturbance. ARIMA (1,0, 1) (0, 1, 1)7 is proposed for the disturbance in a seasonal daily model with a delay factor for the LTF model for Year 2004 which has overlapping moving holidays. Both models recorded 2.075% and 2.606% respectively as the lowest MAPE value and therefore, can be considered for forecasting any public holidays in Malaysia.