In order to solve the problems of inconsistent prediction models,cumbersome calculation process and limited accuracy of prediction results of reinforced concrete beams reinforced with carbon fiber reinforced composite materials(CFRP),this paper combines the two fields of bridge engineering and machine learning,and based on the experimental data of existing literature,the prediction model results under different machine algorithms are constructed and compared,in order to provide a unified calculation model for the prediction of flexural capacity of reinforced concrete beams reinforced with carbon fiber reinforced composites.Help the high-quality development of China's bridge engineering industry.