Abstract—In this paper an approach is proposed for Short Term Load Forecasting (STLF) which combines Wavelet Transform (WT) and Artificial Neural Network (ANN). It is well known that the electrical load at any time can be considered as a linear combination of different frequencies. The daily load curves are decomposed into approximation part associated with low frequency and some detail parts associated with high frequency by means of (WT). Feed Forward Neural Networks are trained by low frequencies and corresponding average temperature or maximum and minimum temperature to predict the approximation part for the next seven days.
The short term load is forecasted by summing the predicted approximation part and the mean of the detail parts of the last three same type days.
This approach had been tested by load of Erbil, the capital of Iraq Kurdistan Region. The result have showed an encouraging accuracy, given that the available data is inaccurate.

Index Terms—Artificial Neural Networks, Short Term Load forecasting, Wavelet Transform.

Cite: A. A. Rasool, A. A. Fttah, and I. B. Sadik, "Short Term Load Forecasting for Erbil Distribution System Using ANN and Wavelet Transform," International Journal of Computer and Electrical Engineering vol. 1, no. 3, pp. 334-339, 2009.