Article ID: | iaor2002513 |
Country: | United Kingdom |
Volume: | 5 |
Issue: | 2 |
Start Page Number: | 131 |
End Page Number: | 141 |
Publication Date: | Jan 1993 |
Journal: | IMA Journal of Mathematics Applied in Business and Industry |
Authors: | Hyde O., Hodnett P.F. |
Keywords: | energy, forecasting: applications |
A short-term load-forecasting system has been developed to predict load demand for the Irish Electricity Supply Board from 1 to 10 days in advance. Model output is acceptably accurate at most times, but special days such as public holidays and the Easter weekend, etc. require individual treatment. The basic model used in everyday load forecasting is presented first, and special days when this model fails in accuracy are then examined and treated separately using rule-based procedures. These procedures are based on identified relationships between weather conditions and prevailing daily load shapes. Comparisons are made between the forecasts for these special days as given by the standard model and the rule-based procedures.