Abstract: It is critical to investigate the key influencing factors of China's carbon emissions in order to achieve the goal of carbon peaking and carbon neutrality. Firstly, a spatial cluster effect analysis of China's provincial carbon emissions in 2003—2019 shows that the spatial distribution characteristics of carbon emissions significantly affect the energy structure of carbon emissions, followed by population size, energy intensity, per capita GDP, urbanization rate andindustrial structure of carbon emissions. Then, as influencing factors of carbon emissions, per capita GDP, energy structure, energy intensity, population size, and urbanization rate were chosen, and a carbon emissions prediction model was developed using an improved bat algorithm and BP neural network. According to the results of the test, the model's average error is 0.16%. Finally, three scenarios of high-, medium-, and low-speed carbon peaking were created for scenario analysis, and the predicted values of carbon emissions after taking into account the carbon sinks were obtained. According to the research, China is expected to reach the carbon peaking target in 2028—2029 under the high- and medium-speed scenarios, corresponding to a carbon emission peak of 12 billion to 12.2 billion tons. This study can provide a reference for the adjustment scheme of energy intensity and energy structure.