Abstract:With regard to the milling process of GH4169, the surface residual stresses under different tool structural parameters were obtained based on orthogonal experimental method. The initial weights and thresholds of the BP neural network were optimized using a genetic algorithm (GA) to improve the convergence speed and prediction accuracy of the model, and a method for applying the GA–BP model to predict the milling residual stress was proposed. The firefly algorithm (FA)–based method for process parameter optimization was investigated, and the GA–BP–FA parameter optimization model for milling residual stresses was established in combination with the GA–BP prediction model for multi-objective optimization of tool geometry parameters with the goal of simultaneously obtaining the minimum residual tensile stress/maximum residual compressive stress. The results show that the minimum residual tensile stress in the X– direction and the maximum residual compressive stress in the Y–direction can be obtained using the optimized tool geometry parameters.