Load forecasting based on wavelet transform and fuzzy logic

Abstract

The electrical loads is the non-stationary time series. With the non-stationary time series, the wavelet technique is more effective technique for approximation. In recent year the fuzzy-wavelet is more attractive for forecasting problem. This paper presented the load curve and electricity consumption forecasting based on MODWT and fuzzy logic. For each individual wavelet sub-series, the fuzzy-rules will be determined in order to carry out the forecasting. After that, the individual wavelet sub-series forecasts are recombined to give the final result.