Stamping Process Design of A-Pillar-Upper Inner Plate Based on PSO-SVR Algorithm

Aiming at the problem of low efficiency and dependence on experience in stamping process design of high strength steel structural parts of automobile body, this paper takes A-pillar-upper inner plate as the object, designs the evaluation index of forming quality and the parametric model of stamping process, forms the evaluation index of twist springback and the parametric model of stamping process, and establishes the support vector machine regression model with stamping process as input and forming quality as output. The particle swarm optimization algorithm is used to optimize the hyperparameters of the SVR model, and then the SVR model coupling stamping process and forming quality is obtained by learning 80 sample data. Finally, according to the expected forming quality of A-pillar-upper inner plate, the PSO algorithm is used to solve the SVR model in reverse, and the parameters of the stamping process are obtained. After the trial production of the mold and parts, the measured twist springback of the parts is 0.106, while the predicted value of the model is 0.101, and the relative error is only 5 %, which proves the reliability of the process design using this method.