基于6σ的多学科设计优化

在实际的工程问题中,总是不可避免地遇到一些不确定因素,导致许多优化策略无法用数学模型来体现这些不确定因素,从而造成结果分析错误或不全面．因此引入面向6σ概率设计(DFSS,σ为随机变量的标准偏差)方法,对其算法原理进行了阐述和分析,并且采用多学科设计优化(MDO)高层优化算法框架来支持DFSS设计,从而可以有效避免DFSS在处理复杂问题时串行处理的缺点,将涉及多学科的复杂系统问题逐步分解为各个单学科来并行地进行概率设计．最后,在理论研究的基础上结合轴承座这个工程实例,给出了其具体实现,证明了将二者结合进行复杂问题的概率设计是可行的．

Multidisciplinary design optimization based on DFSS

Some uncertainty factors are inevitable in practical engineering problems, which would result in many optimization strategies not being expressed by mathematical model,and would lead to errors in analysis. This paper introduces methods of Design for Six Sigma(DFSS) and its theories,and originally adopts Multidis-ciplinary Optimization Design(MDO) to support DFSS, which can avoid its serial process when to meet with complicated problems and divide complicated ones into individuals. The connection of the two proves to be feasible for complicated problems by the case of bearing base.