Short term load forecasting using artificial neural network with feature extraction method and stationary output

This paper presents the artificial neural network (ANN) that used to perform STLF for the next 24 hours. The feature extraction involves a transformation of raw data that is from the chronological hourly peak loads to the multiple time lags of hourly peak loads. This is used to improve the input data which will significantly enhance the performance of ANN in forecasting the hourly peak loads with less error. The output of ANN is then converted to a non-stationary form which represents as the forecasted hourly peak load for the next 24 hour. The Malaysian hourly peak loads in the year 2002 is used as case study to verify the effectiveness of ANN in STLF.