Abstract：This study takes Guangdong Province as an example to establish three different carbon emission prediction models for predicting cement clinker and cement production, and the prediction accuracy of different methods is compared. Next, this study sets benchmark scenarios, policy scenarios and enhanced scenarios to analyze the development trend of carbon emissions in the cement industry, and develops corresponding carbon peak paths and policy recommendations. The research results indicated that the prediction accuracy of genetic algorithm optimized BP neural network model for carbon emissions was higher than that of multiple linear regression analysis method and support vector machine method. Based on the research results of scenario analysis method, it can be seen that the carbon emissions of the cement industry in Guangdong Province reached a peak in the early stage of the 14th Five Year Plan, which has a peak of approximately 128-130 million tons. During the middle and later stages of the 14th Five Year Plan and the 15th Five Year Plan, there was a continuous downward trend. The carbon reduction measures with the potention from large to small in the cement industry are fuel substitution, energy-saving transformation and raw material substitution. The cement industry in Guangdong Province can achieve carbon peak through industrial structure path, energy structure path, outdated production capacity elimination path, and energy-saving and carbon reduction technology path
2023, Vol. 0 
