Modeling and Forecasting Energy Intensity in China by Using The Four-parameter Logistic Model and An Artificial Neural Network

Abstract

ABSTRACT The main goal of this study was to model the evolution of energy intensity in China by using the four-parameter logistic model and the artificial neural network technique to determine the future level of energy intensity in China. In this study, a back-propagation network with one hidden layer was used. The two models were compared for degree of fit to the historical data on energy intensity in China for the period of 1980 to 2009. The selected statistical measures include SSE, RMSE, MAPE, R2 and residual plots. The comparison indicates that the better modelling of the evolution of energy intensity in China is given by the developed artificial neural network model. According to the results forecast by it, energy intensity in China in the future will slowly decrease.