Forecasting petroleum consumption in China: comparison of three models

There have been a number of forecasting models based on various sigmoidal growth models. This paper investigates the effectiveness of the four-parameter logistic model, the Gompertz-Laird model and the stochastic Gompertz innovation diffusion model for describing the evolution of petroleum consumption in China. The three sigmoidal growth models are applied to the historical data on petroleum consumption in China. The developed models are compared using the goodness-of-fit to the historical data. The selected statistical measures are the coefficient of determination R-2, the mean squared error, the mean absolute percentage error, the mean absolute deviation and the Durbin-Watson statistic d. The good model fit has indicated that the four-parameter logistic model is a very appropriate candidate in forecasting petroleum consumption in China.