Accurate forecasts and comparative analysis of Chinese CO2 emissions using a superior time-delay grey model

In China's new development stage, reaching carbon peaking and neutrality has emerged as a complicated and substantial task, highlighting the importance of forecasting CO2 emissions. Therefore, this research designs a novel convolution-based multivariate time-delay grey model that integrates time-delay coefficients and analytical solutions, thus accurately identifying and quantifying time-delay effects. Moreover, we present an algorithm determination framework to enhance the model's applicability and flexibility, and assess the robustness through the Monte Carlo Simulation and parameter sensitivity analysis. For demonstration purposes, the novel technique is utilized to forecast Chinese CO2 emissions, yielding the lowest MAPE values (1.80%for three-steps ahead, 1.51% for two-steps ahead, and 0.27% for one-step ahead) compared with traditional grey and non-grey competitor models. Additionally, the Monte Carlo Simulation results reveal that the particle swarm optimization algorithm outperforms other candidate algorithms due to the highest accuracy and optimal stability. Overall, the newly-proposed approach, possessing more powerful flexibility and applicability than the previous models, is an effective technique for forecasting CO2 emissions in China.