A novel nonlinear multivariable Verhulst grey prediction model: A case study of oil consumption forecasting in China

Oil resources affect the development of the global economy, so forecasting oil consumption is a necessary basis for formulating economic and social plans. In this paper, the characteristics of all background values in a system are considered, and a genetic optimization algorithm is used to establish a new nonlinear multivariable Verhulst model. This model weakens the demand of the Verhulst model for saturated S-shaped and single-peaked data, thus increasing its applicability. To verify the validity of the novel extended model, eight evaluation indices are utilized in actual cases. The outcomes reveal that the proposed model significantly outperforms the preoptimized grey multivariate Verhulst model and other traditional grey models. Finally, the proposed model is employed for the prediction of oil consumption in China, and a comparison is made with nine models, including a neural network model, ARIMA, a grey linear model and a grey nonlinear model. The findings of a comparison of eight evaluation metrics show that the new model is second only to the neural network model in prediction performance, with the gap being small. The new model predicts that China’s oil consumption will increase by 24.6641% in 2024. This forecasted information can provide a reference for relevant units and individuals in China and the global oil market.