基于参数和代理模型不确定性的冲压稳健性设计优化

Robust Design Optimization for Stamping Based on Parametric and Metamodel Uncertainty

在传统的基于代理模型的冲压稳健性设计中,由于代理模型与真实模型间存在着误差,必然会导致优化结果存在一定的误差。将实验设计、Kriging模型相结合,综合考虑参数不确定性和代理模型不确定性的影响,提出了一种新的冲压稳健性优化设计方法。通过因素敏感性分析筛选出相应的设计变量和噪声因素,基于Kriging模型构建设计参数和质量指标的代理模型,采用蒙特卡罗分析方法以及遗传算法获得最优工艺解。实例分析结果表明,综合两种不确定因素的稳健设计方法能有效地降低拉裂、起皱约束失效概率,提高冲压件成形质量和工艺稳健性。

In traditional metamodel-based stamping robust design,due to the errors between meta-model and real model,there were certain design errors in the optimization results.Combining experi-mental design with Kriging model and considering both parametric and metamodel uncertainty,a new robust design method was established herein.The agent models of design parameters and quality index were built,and the Monte Carlo simulation method was combined with genetic optimization algorithm to obtain the optimal process solutions.The results show that compared to the traditional design method,the proposed method can effectively reduce the probability of crack and wrinkle constraint failure,and improve the forming quality and process robustness.