Purpose
The purpose of this paper is to compare GM(1,1) model, rolling GM(1,1) model and metabolism GM(1,1) model included in the GM(1,1) model cluster and use these models to forecast global oil consumption.
Design/methodology/approach
Simulated sequences will be generated randomly, and used to test the models included in the GM(1,1) model cluster; and these grey forecasting models are applied to forecast global oil consumption.
Findings
Effectiveness of these grey forecasting models is proved by random experiments, which explains the model adaptability. Global oil consumption is predicted, and it shows that global oil consumption will increase at a rather big growth rate in the next years.
Originality/value
The effectiveness of medium-term prediction of these grey forecasting models is analyzed by random experiments. These models are compared, and some basis for model selection is obtained.