基于灰色关联度的动态稳健性设计  

Grey relation grade for the analysis of robust designs with dynamic characteristics

作　　者：钟晓芳[1,2] 刘思峰[1] 

机构地区：[1]南京航空航天大学经济与管理学院管理科学与工程系,南京210016 [2]南京理工大学经济管理学院管理科学与工程系,南京210094

出　　处：《系统工程理论与实践》2009年第9期147-152,共6页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70701017);江苏省博士后科研基金(0802065C);江苏省高等学校优秀创新团队(P0702);南京航空航天大学创新群体(Y0553)

摘　　要：针对动态特性稳健性设计,摒弃传统的信噪比分析方法,直接由响应变量与信号因子之间的线性关系,利用灰色关联度衡量响应变量与信号因子之间的关联程度,选取一组使得关联度最大的可控因子水平组合;同时分析不同使用条件下的波动与零向量之间的关联程度,同样选取一组使得关联度最大,即波动最小的可控因子水平组合.然后对二者进行综合比较分析,选取一组使得二者均能达到满意结果的可控因子组合,并进行验证.最后以一个案例加以说明,得到了比传统信噪比方法更好的结果,并指出了进一步研究的方向.In view of robust design with dynamic characteristics, traditional analysis method with signal to noise ratio has been given up in this paper. A new grey relation grade analysis has been used to measure the correlation degree between the response variable and the signal factor, so as to select a group of control settings which make response performance well over different input conditions. Simultaneously the correlation degree of the fluctuation under the different conditions and zero has been analyzed and another group of similar combination has been selected. Then the two groups are compared and analyzed to choose the one which makes a system＇s performance robust over a wide range of input conditions. Finally, a case is illustrated to achieve the better result instead of traditional signal to noise ratio method, and the further study has also been introduced.

关 键 词：灰色关联度 稳健性设计 动态特性 响应变量 信号因子 可控因子 

分 类 号：TB114[理学—概率论与数理统计]