Long term electric load demand forecasting of Turkey

The importance of electrical energy in the age of technology now is increasing day by day. The utilities, which are responsible for the production; transmission and distribution of the electricity, are trying to provide the electricity to the customers as possible as continuous; high quality, and inexpensive. In order to fulfill the requirements, it has been inevitable that they use the past data related to the usage of electricity and make future consumption projections. Of course, there are wide range of methods that can be used to make forecasts. In this study, long-term electric load demand forecasting of Turkey has been performed by using time series analysis; multiple linear regression and artificial neural networks (ANN). The results of the methods are compared with each others? and with the results of MAED program which is used by the Energy and Natural Resources Ministry of Turkey. In the study, it is determined that the artificial neural network could forecast the electricity demand for the past years with less error rate compared to the other two methods. In addition, the future projections of the ANN method are found to be close to the projections of the MAED method.