动态多响应系统的稳健参数设计

Robust parameter design for dynamic multi-response system

针对动态多响应之间的相关性及试验设计的信息不确定性,提出了一种动态多响应系统的稳健参数设计方法.首先,利用满意度函数对动态系统的敏感度和波动分别进行了度量,在此基础上进行了主成分分析;其次,采用灰色关联分析得到主成分序列与其理想序列的灰色关联度,然后通过TOPSIS方法转换得到综合灰色关联度,并根据综合灰色关联度确定可控因子的最优水平组合;最后,通过两个实例说明了该方法的有效性与优越性.

In view of the correlations among multiple responses and information uncertainty of design of experiment （DOE）, a new method of robust parameter design for dynamic multi-response system was proposed. Firstly, desirability function was used to measure dynamic system sensitivity and variation, and principal component analyses on the two indices were conducted. Secondly, grey relational analysis was used to gain the grey relational grade （GRD） of principal component sequences relative to their respective ideal sequences, and transferred them into integrated grey relational grad （IGRD） by means of TOPSIS, then identified the optimal level combination of control factors based on the IGRD. Finally, the validity and superiority of the proposed method are confirmed through two case studies.