ABSTRACT

In this paper, a new grey Theta forecasting model is established to predict primary energy consumption. The parameter θ is used to adjust the slope of this trend. In addition, this hybrid method can be used in combination with other forms of grey model, which has great potential to improve the accuracy of prediction. The pseudo-code of the model is given, and the computational complexity of the proposed method is calculated. The robustness of the model is measured by the examples of primary energy consumption in China, the United States, India, Russian Federation and Japan. Numerical examples show that the proposed model is superior to other competitive models. Based on the local curvature adjustment of the Theta line, the proposed method can improve the predictive performance of the grey model in nonlinear trend data. Then, the future trends of global primary energy consumption and total carbon dioxide emissions are also forecasted. Predictions show that only China and India are expected to see annual primary energy demand rise by more than 5% over the next five years, while the United States, Japan and other countries will see no growth in demand for primary energy.