Forecasting CO2 emissions using a novel fractional discrete grey Bernoulli model: A case of Shaanxi in China

Abstract

To accurately predict CO2 emissions, this paper proposes a novel fractional accumulation method and constructs a novel fractional discrete grey Bernoulli model DGFGBM (1,1,tα). The manta ray foraging optimization (MRFO) algorithm is used to find the parameters of the new model, and the carbon emission data of Shaanxi Province in China and three of its cities are applied to verify the effectiveness of the new model. The CO2 emissions of the four regions from 2020 to 2024 are predicted, and the following conclusions can be drawn. The new fractional accumulation method is effective and reasonable and can make full use of new information and improve the prediction accuracy of the model. Compared with three other optimization algorithms, MRFO has more advantages in finding the optimal parameters of the model. Compared with other grey models and the ARIMA model, the DGFGBM (1,1,tα) model has better prediction performance. The prediction results show that from 2020 to 2024, the CO2 emissions of Shaanxi Province are expected to increase by 94.6278 million tons, an increase of 31.8%, and the CO2 emissions of Xi'an, Xianyang and Baoji are expected to increase by 22.4901 million tons, 16.3129 million tons and 7.3271 million tons, respectively, with growth rates of 45.81%, 46.51% and 34.07%, respectively.