Abstract: A method combining back propagation neural network (BP neural network) and genetic algorithm was proposed in this paper in order to improve the warping uniformity of air-conditioner blades. Mold temperature, melt temperature, injection time, packing time, and packing pressure were taken as design variables and Z axis maximum difference in the blade tip was as optimization goal. After that, CAE simulation was conducted based on Taguchi method. A BP neural network model was developed to obtain the mathematical relationship between the optimization goal and design variables, and genetic algorithm was applied to optimize the process parameters. Consequently, the optimal process parameters were obtained as: mold temperature 45 ℃，melt temperature 205 ℃，injection time 1.8 s，packing time 6 s，packing pressure 50 MPa. Finally，the CAE simulation was simulated under the optimized parameters. As a result, the target variable was 0.08 mm, lower than the experimental scheme, and the blade warping uniformity was improved.