车身零件定位方案的多变量稳健设计  

Multivariable Robust Locating Scheme Design of Body Parts

作　　者：邹景明[1] 储国平[1] 梁树勇 尹国丽 

机构地区：[1]上海交通大学上海市复杂薄板结构数字化制造重点实验室 [2]上汽通用五菱汽车股份有限公司技术中心

出　　处：《上海交通大学学报》2013年第9期1353-1357,共5页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金委员会创新研究群体科学基金(51121063);教育部高等学校学科创新引智计划(B06012)资助

摘　　要：在新车型初始设计阶段仅有零件的A面数据或初步结构为已知的情况下,提出了一种在初始设计阶段零件装配的定位方案稳健设计方法,建立了基于刚性零件确定性定位和稳健设计理论的定位方案优化设计模型,使用遗传算法在可行定义域内求解最优定位点的位置,并提取了定位点偏差对整体装配效果的敏感度系数,实现了在优化公差的同时控制生产成本.同时,以某新车型前大灯的定位方案设计为例,验证了所提方法在初始设计阶段的有效性.A robust locating scheme design was provided under the circumstance that only parts of A sur- face or initial structure are known in the initial design phase of a new vehicle model. An optimization de- sign model of locating scheme was established based on the deterministic locating for rigid parts and robust design principles. The optimal locators are searched by using genetic algorithms in feasible region; the sen- sitivity coefficient which provides an indicator of how sensitive the assembly quality are to variance of loca- tor is extracted, and finally tolerance optimization is achieved while costs are reduced. An example of loca- ting design of a new vehicle model＇s head-lamp was demonstrated, which proves that the method is effec- tive in the initial design phase.

关 键 词：定位设计 可行定义域 偏差 敏感度系数 

分 类 号：TH161[机械工程—机械制造及自动化]