Prediction and Comparative Analysis of Peak Stress Using Optimized Constitutive and ATPSO-BP Models: A Case Study of 6016H18 Aluminum Alloy

In order to increase the prediction accuracy of the peak of flow stress for the 6016H18 aluminum alloy, the Browman constitutive model is optimized by quadratic polynomial fitting instead of linear fitting, and back propagation (BP) is improved by active target particle swarm optimization (ATPSO) to obtain the ATPSO-BP model. The peak stress of the 6016H18 aluminum alloy is predicted and comparatively analyzed. According to the numerical results, the prediction precision of the optimized Browman constitutive model is higher than that of the original one. It is confirmed by the mean relative error that the ATPSO-BP model performs better in terms of prediction than the Browman constitutive model, the optimized model, and the BP model, which can objectively reflect the influence of all heat deformation parameters on peak stress. The related results can provide an important theoretical basis for studying the mechanical properties of aluminum alloy and provide a technical reference for the application of the 6016H18 aluminum alloy.