Application of the decomposition technique for forecasting the load of a large electric power network

A reliable electric load forecasting technique for power system planning is introduced. In this technique, a suitable model to fit the data is selected before the actual forecasting algorithm is performed. The model is constructed in three steps: identification, estimation and diagnostic checking. This model is then used by two decomposition techniques (multiplicative and additive) to forecast the electric load. The techniques are illustrated by applying them to the forecast of the monthly peak loads of a large electric power network. The Egyptian electric power network peak loads data are used in this study. Analysis and comparison of the results have shown that the techniques are adequate and yield good results.