A neuro wavelet-based approach for short-term load forecasting in integrated generation systems

In the paper is proposed a new neuro-wavelet based approach for the problem of short term load forecasting. The implemented neuro-wavelet based algorithm combines the potential of two soft computing techniques. The strength over other approaches appeared in literature is that firstly the hourly power load data are wavelet processed and then provided as input to an RNN. The obtained simulation results confirm the improved forecasting model over conventional techniques.