This research is concerned with the effects of the geometrical parameters of the die in elevated temperature Hydro-Mechanical Deep Drawing (HMDD) process of 2024 aluminum alloy. A Group Method of Data Handling (GMDH) process was used to train a neural network in order to study the process behavior. Based on the maximum reduction in sheet thickness and the uniformity of the final product, an objective function was constructed. The Bees Algorithm (BA) was used to achieve the optimal values for process variables. To verify the simulation results, they were compared with the experimental findings gained via this research and an appropriate correlation was observed between these results. This comparison showed that, by optimization of the geometrical parameters of the process, the value of the combined objective function was the best one compared with all of the cases tried in the present investigation.

这项研究与模具几何参数对2024铝合金高温液压机械深冲（HMDD）工艺的影响有关。为了研究过程行为，使用了分组数据处理（GMDH）过程来训练神经网络。基于最大的片材厚度减少和最终产品的均匀性，构建了目标函数。Bees算法（BA）用于获得过程变量的最佳值。为了验证模拟结果，将它们与通过这项研究获得的实验结果进行了比较，并观察到了这些结果之间的适当相关性。比较结果表明，通过优化流程的几何参数，