As the accuracy of the electricity load forecast is crucial in providing better cost-effective risk management plans, this\npaper proposes a Short-Term Electricity Load Forecast (STLF) model with a high forecasting accuracy. A cascaded forward BPN\nneuro-wavelet forecast model was adapted to perform the STLF. The model was composed of several neural networks whose\ndata were processed using a wavelet technique. The data used in the model was electricity load historical data. The historical\nelectricity load data was decomposed into several wavelet coefficients using the Discrete wavelet transform (DWT). The wavelet\ncoefficients were used to train the neural networks (NNs) and later, used as the inputs to the NNs for electricity load prediction.\nThe Levenberg-Marquardt (LM) algorithm was selected as the training algorithm for the NNs. To obtain the final forecast, the\noutputs from the NNs were recombined using the same wavelet technique.
Loading....