An MPA-based optimized grey Bernoulli model for China's petroleum consumption forecasting

The remarkable prediction of petroleum consumption is of significance for energy scheduling and economic development. Considering the uncertainty and volatility of petroleum system, this paper presents a nonlinear grey Bernoulli model with combined fractional accumulated generation operator to forecast China's petroleum consumption and terminal consumption. The newly designed model introduces a combined fractional accumulated generation operator by incorporating the traditional fractional accumulation and conformable fractional accumulation; compared to the old accumulation, the newly optimized accumulation can enhance flexible ability to excavate the development patterns of time-series. In addition, to further improve the prediction performance of the new model, marine predation algorithm is applied to determine the optimal emerging coefficients such as fractional accumulation order. Furthermore, the proposed model is verified by a numerical example of coal consumption; and this newly established model is applied to predict China's petroleum consumption and terminal consumption. Our tests suggest that the designed ONGBM(1,1,k,c) model outperforms the other benchmark models. Finally, we predict China's petroleum consumption in the following years with the aid of the optimized model. According to the forecasts of this paper, some suggestions are provided for policy-makers in the relevant sectors.

Automated summary

This paper proposes a nonlinear grey Bernoulli model with a combined fractional accumulated generation operator for forecasting China's petroleum consumption. The model is optimized using a marine predation algorithm. The results show that the proposed model outperforms other benchmark models in terms of prediction accuracy.