A Process design system for vehicle stamping parts based on SVR-PSO

Abstract

Aiming at the problem that the stamping process design of automobile structural parts is not standardized and overly dependent on experience, this paper designs an intelligent design system of the stamping process based on a support vector machine. The system includes three modules: management data, training model, and design process. The system can save a large amount of stamping process design data in a standardized form in the system, and then obtain a support vector machine regression model. When designing the process, the user only needs to input the geometric parameters, material type, and expected forming quality of the part, and the system will solve the regression model based on the constraint conditions through the PSO algorithm to obtain the best process parameters. In practical application, taking the A-pillar-upper as an example, the measured springback of the part is 2.73, while the predicted value of the model is 3, and the relative error is only 9%, which proves the reliability of the process design using this method.