Combining wavelet transform and support vector regression model for day-ahead peak load forecasting in the Greek power system

Day-ahead peak load forecasting is an essential tool for generation units scheduling, unit commitment and generally, in power systems operation in short-term horizon. The scope of the present study is to develop a robust peak load forecasting model for the power system of Greece. The peak load series is decomposed via the Discrete Wavelet Transform (DWT) into low-frequency and high-frequency subseries in the wavelet domain. For each subseries the Support Vector Regression (SVR) model is trained and applied. The final peak load series is obtained by the inverse DWT. The proposed approach corresponds to low execution time and robust performance and thus, can be a valuable tool for utilities, systems operators and others.