The purpose of this paper is to investigate the effectiveness of GM(1,1) model on linear growth sequences (LGS) by random experiments and global primary energy consumption is predicted as by the GM(1,1) and the autoregressive integrated moving average (ARIMA) model, which is used as a reference.
LGS generated randomly are used for GM(1,1) modeling. The results of the massive repeated random experiments are analyzed to test the effectiveness of the GM(1,1) model and global primary energy consumption is predicted using the GM(1,1) model and the ARIMA model.
The use of the GM(1,1) model is effective when used for a LGS and the model is proven to be reliable by the experiments. Global primary energy consumption is predicted with the GM(1,1) model and the ARIMA model as a case study, and the results show that GM(1,1) is quite good. Global primary energy consumption will increase by 1.03 percent in 2016.
The contribution of this paper includes the following: first, the applicability of the GM (1,1) model is further discussed with random experiments and it is feasible for a LGS; second, random experiments provide good proof that four data are enough for GM(1,1) modeling, and GM(1,1) model is reliable; third, prediction by using GM(1,1) model with small data is even better than time-series analysis in the case study.
