Abstract: Raymond mill is widely used in the production and processing of building materials, but the turbine of its classifier has the problem of unsmooth airflow, which seriously affects its working efficiency. In order to solve this problem. Firstly, the model was established and the fluid simulation analysis was carried out. In Solidworks, the parametric modeling of the Turbine of Raymond mill classifier was carried out. This model was mainly based on the top-down design of Solidworks 2020 to complete the correlation between parameters. Then the original aerodynamic drag coefficient was obtained by aerodynamic analysis with finite element software. Then, Isight software combined with Solidworks, Icem and Fluent was used to optimize the turbine structure of Raymond mill classifier. Four design variables were selected to conduct DOE test design in Isight, and a second-order response surface approximation model was established to analyze the error of the optimization model, and the goodness of fit is 0.98, so as to confirm the feasibility of the model. Finally, combined with sequential quadratic programming (NLPQL) and the minimum aerodynamic drag coefficient as the objective function, the structural parameters were optimized and calculated. Finally, the aerodynamic resistance coefficient in Z-axis direction of Raymond mill is reduced from 21.44 before optimization to 18.97 after optimization, and the aerodynamic resistance coefficient in Z-axis direction is reduced by 11.52%, showing a good effect.