DYNAMIC MODELLING AND FORECASTING OF MONTHLY ELECTRIC ENERGY DEMAND IN THE HUNGARIAN ELECTRIC ENERGY SYSTEM

Abstract

This paper discusses the model building for forecasting Electric Energy Demand /EED/ in the Hungarian Power System. Based on the regularities observed in the monthly EED, some hypothetical model structures are derived, that suggest the use of SARIMA time series models. The effect of temperature can also be treated using transfer function models. By the application of adaptive parameter estimation methods, transfer function model having time varying parameters /due to time dependent temperature sensitivity/ can also be identified.