Applications of SVM and improved particle swarm algorithm to sheet metal forming optimization

In view of the complex variable relation in the optimization of sheet metal forming process, a novel optimization method is proposed. By applying three strategies (nonlinear dynamic improvement of inertial weights, nonlinear dynamic adjustment of acceleration factor and introduction of adaptive particle mutation) to the improvement of standard particle swarm optimization algorithm, an accurate regression model between process parameters and forming quality is constructed based on support vector machine, and combined with improved particle swarm optimization algorithm, an optimization on the parameters of sheet metal forming is conducted. The results of optimization effectively control the cracking and wrinkling defects and improve the quality of sheet metal forming. ©, 2015, SAE-China. All right reserved.