<p>Global climate deterioration is increasingly severe, and the carbon emission problem has sounded the alarm for the survival of humanity. Scientific carbon emission forecasting is the core of formulating the emission reduction strategy. This paper forecasts the total carbon emissions in China with a multivariable grey model considering the nonlinear interaction between related factors. It improves data accumulation and introduces the nonlinear interaction of related factors to optimize the traditional modeling mechanism. The fitting and test accuracies of the proposed grey model are 0.38% and 0.18%. Based on the forecast, China's carbon emissions may reach 13815.56 MtCO₂, and the carbon emission intensity may continue to decline. The forecast framework proposed in this paper deepens the discussion of the future development trend of China's carbon emissions from consuming three primary fossil fuels. The results provide quantitative references for reducing carbon emissions and emission intensity targets. It reveals that if the annual growth rate of coal consumption among the three primary fossil fuels is controlled below 0.5%, China could achieve its carbon peak target before 2030. This paper indicates that the Chinese government should reduce its reliance on coal resources and design a scientific coal control road for dual-carbon goals.</p>