摘要
为了提高冲压件的可靠性和稳定性,减少代理模型不确定性和参数不确定性对冲压产品质量的影响,需要进行稳健设计。提出了一种具有内外层结构的容差稳健模型,有效地降低了可控因素和不可控因素不确定性引起的质量波动。通过似然函数因素筛选法确定关键的可控因素和不可控因素,采用改进的贝叶斯估计对代理模型不确定性进行量化。以NUMISHEET93方盒件为研究对象进行稳健优化设计,利用目标函数和约束函数的均值、方差建立容差稳健模型。将稳健优化结果与忽略模型不确定性的优化结果分别进行蒙特卡洛模拟并对比。结果表明,综合考虑代理模型不确定性和参数不确定性的稳健设计方法能有效地提高冲压件的成形质量。
In order to increase the reliability and stability of stamping parts and reduce the influences of metamodel uncertainty and parameter uncertainty on the quality of stamping parts,a robust design is needed.Then,a tolerance robust model with inner and outer loop structures was proposed,and the quality fluctuation caused by the uncertainty of controllable factors and uncontrollable factors was reduced effectively.Furthermore,the key controllable factors and uncontrollable factors were confirmed by likelihood function factor selection method,and the metamodel uncertainty was quantitated by revised Bayesian estimation.For the square box of NUMISHEET 93,the robustness optimization design was conducted,and the tolerance robust model was established by the mean and variance of objective and constraint functions.In addition,the robust optimization results were compared with the optimization results of ignoring metamodel uncertainty by Monte Carlo simulation.The results show that the robust design method considering comprehensively metamodel uncertainty and parameter uncertainty can effectively improve the forming quality of stamping parts.
作者
潘贝贝
谢延敏
岳跃鹏
冯美强
张飞
Pan Beibei;Xie Yanmin;Yue Yaopeng;Feng Meiqiang;Zhang Fei(Institute of Advanced Design and Manufacturing,Southwest Jiaotong University,Chengdu 610031,China)
出处
《锻压技术》
CAS
CSCD
北大核心
2019年第4期182-187,共6页
Forging & Stamping Technology
基金
国家自然科学基金资助项目(51005193)
四川省科技计划项目(2019YFG0313)
关键词
冲压件
代理模型不确定性
参数不确定性
容差稳健模型
似然函数因素筛选法
蒙特卡洛模拟
stamping part
metamodel uncertainty
parameter uncertainty
tolerance robust model
likelihood function factor selection method
Monte Carlo simulation
