Abstract: In order to analyze whether China can achieve the goal of carbon peak on time, this paper collects 23 potential influencing factors of carbon emission from the dimensions of population, economy and energy, and constructs a multi-scenario prediction model based on the Pelican Optimization-Hybrid Kernel Extreme Learning Machine through factor screening. This paper sets five scenarios based on different economic development modes, and further analyzes the patterns of achieving carbon peaks between different paths using the tapio decoupling model. The results show that China can achieve carbon peak on time under the scenario of industrial structure adjustment, ecological development and overall development; in the baseline scenario, the carbon peak will arrive in 2032, and if the Chinese government allows industrial economic development without strengthening environmental protection, the carbon peak will not be achieved until 2033; and the earlier the carbon peak is achieved, the stronger the decoupling effect between carbon emissions and economic growth will be.