Power system load forecasting based on EEMD and ANN

Abstract

In order to fully mine the characteristics of load data and improve the accuracy of power system load forecasting, a load forecasting model based on Ensemble Empirical Mode Decomposition (EEMD) and Artificial Neural Networks (ANN) is proposed in this paper. Firstly, the load data can be resolved into a limited number of Intrinsic Mode Function (IMF) components and one remainder by EEMD which avoids the mode mixing problem of Empirical Mode Decomposition (EMD). Then, through the observation of the spectrum by Hilbert transform, it's obvious that the regularity and periodicity of low frequency components are stronger than high frequency components. So one sole appropriate ANN forecasting model is chosen for each low frequency component, and the linear combination of ANN model is applied to forecasting each high frequency component. Simulation results show that the new model proposed in paper is better than anyone ANN forecasting model. © 2011 Springer-Verlag.