웨이브릿 변환을 이용한 계절별 부하예측 알고리즘

Seasonal load forecasting algorithm using wavelet transform analysis

This paper proposes a novel wavelet transform based algorithm for the seasonal load forecasting. In this paper, Daubechies DB2, DB4 and DB10 wavelet transforms are adopted to predict the seasonal loads and the numerical results reveal that certain wavelet components can effectively be used to identify the load characteristics in electric power systems. The wavelet coefficients associated with certain frequency and time localization are adjusted using the conventional multiple regression method and then reconstructed. In order to forecast the final loads through a four-scale synthesis technique. The outcome of the study clearly indicates that the wavelet transform approach can be used as an attractive and effective means of the seasonal load forecasting.