Application of soft computing techniques to forecast monthly electricity demand

Abstract

Electricity demand forecasting is an important tool for private enterprise to develop electricity supply system. The purpose of this study is to develop monthly electricity forecasting model in order to predict future electricity demand for energy management. The influence of the weather factors such as temperature and humidity are diluted in an overall value that represents the total monthly electricity demand. So, the forecasting model uses only historical electricity demand data to obtain future prediction. This study presents an approach to monthly electricity demand time series forecasting model, including two series of the fluctuation and trend series. The fluctuation series describe the trend of the electricity demand series and the fluctuation series describe the periodic fluctuation that imbedded in the trend. Then an integrated genetic algorithm and neural network model are trained for forecasting purposes. In order to verify the model, an empirical study was conducted in a private enterprise. Validation is made by comparing with model that only neural network was used.