Application Wavelet for Middle Term Demand Forecasting

Middle term demand forecasting is necessary problem for power distribution planning. For some reasons, some times the forecasted values such as GDP, climate factors and other social-economic index with the acceptable accuracy that are used for multi regression forecasting model are not available. So the forecasting model has to be build basing on the time series of annual electric consumptions. In this paper we tried to use the wavelet technique after analyzing the data. The application of wavelet is effective for forecasting if the origin data are transformed into necessary form. The influence of number of wavelet and embedded dimension is also carried out in this paper