Design of process parameters using robust design techniques and multiple criteria optimization
A. Song, A. Mathur, K.R. Pattipati, S. Deb · 2002
Presents a methodology for the design of products/processes that makes use of the concepts of robust design and the techniques of multiple criteria optimization for simultaneously optimizing many design criteria. First, a systematic approach to the selection of an efficient matrix experiment for a design problem is presented. Appropriate performance measures are obtained so that their joint optimization results in the minimum variation of product quality. A two level discrete multiple criteria optimization algorithm that incorporates the methods of dominated approximations and reference points is developed to obtain nondominated solutions for the design problem. The methodology is illustrated using a case study.>