Methodology for weekly load forecasting

In this paper, a new modeling approach for long range weekly load forecasting is described. Performance of the proposed methodology is verified with simulations of actual data pertaining to the Riyadh fast developing power system operating in the central part of Saudi Arabia. Validity of the methodology is demonstrated by comparing the forecasts obtained from this approach with corresponding ones produced by established time series methods. Sensitivity of the demand forecast to changes in some variables is also discussed.