In response to the urgent global challenge of carbon emission reduction, reliable pathways to carbon peaking are of significant importance for the implementation of carbon reduction in China. However, the process of carbon dioxide emissions involves numerous influencing factors, and the combinations of their change rates are diverse. Traditional scenario analysis methods only enumerate a limited number of possible scenarios and struggle to select the optimal combination of change rates. To address this, this paper takes Fujian Province as an example, and constructs an SSA-SVR (Sparrow Search Algorithm-Support Vector Regression) model based on the analysis of Fujian′s energy consumption and carbon emission data. The model comprehensively considers 14 key factors affecting carbon emissions and uses the SVR model to predict and verify Fujian′s carbon emissions from 1999 to 2022, ensuring the accuracy and reliability of the model. By optimizing the annual change rate combinations of these factors through the SSA algorithm, the study seeks to find possible pathways to meet the carbon peak target by 2030. The research finds that although there are significant differences in the carbon emissions of all explored pathways, they can all achieve a carbon peak by 2030. The experimental results also show that the SSA -SVR model can explore a variety of effective carbon peak pathways, providing policy makers with a diverse selection of emission reduction pathways and offering a scientific basis and strategic recommendations for the industrial sector of Fujian Province to achieve the carbon peak target.