Multi-Objective Optimization of Vortex Disc Process Parameters Based on Neural Network and Genetic Algorithm

In this paper, a multi-objective optimization strategy of production process parameters based on the neural network and genetic algorithm is proposed with an automotive scroll disk as the research object. The forging forming process is numerically simulated by Deform-3D finite element software, with billet temperature, die temperature, and forming speed as optimization variables, and forming load, residual stress, and die deformation as optimization indicators. The nonlinear mapping relationship between variables and indicators is constructed by using the neural network, and the neural network model is optimized based on the genetic algorithm for dynamic optimization of parameters. The most suitable solutions finally obtained in the Pareto frontier set: billet temperature: 460°C, mold temperature: 220.006°C, forming speed: 18.4158 mm/s, when the values of the three optimized indicators are smaller. The solution was experimentally verified and the obtained vortex discs were filled to the brim with no defects, so the process parameters can be applied to actual production processing.