![]() | يوجد فقط 14 صفحة متاحة للعرض العام |
المستخلص This research aims to build a model of FARIMA (p, d, q) or FARIMAX (p, d, q) by applying to the amount of electricity consumed and comparison between the two models. In an attempt to build a FARIMA (p, d, q) model for forecasting the consumed quantity of electricity in residential and industrial sectors as they both consume about 73% of the total consumed quantity of electricity in Egypt. Using time series data from July 1999 to June 2007 in the estimation stage and another 12 monthly observations from July 2007 to June 2008 have been used to test the estimated model. The statistical results concerning the residential sector showed that, the most important independent variables that are affecting the consumed quantity of electricity are the number of monthly consumers and the average price of kilowatt-hour per month. The appropriate model for the data is FARIMAX (0, d, 12, X) with value of fractional differences (d = 0.371), using maximum likelihood in the estimation of the final model. For the industrial sector, it is found that the most important independent variables that are affecting the consumed quantity of electricity are the number of monthly consumers and average price of kilowatt-hour per month. The appropriate model for the data is FARIMAX (2, d, 36, X) with value of fractional differences (d = 0.492), using maximum likelihood in the estimation of the final model. |