POSSIBILISTIC MULTIRESPONSE QUALITY DESIGN

Multi-response quality design techniques are used to identify settings of process parameters that make a product's performance close to a ta rget value in the presence of multiple quality characteristics. In man y situations, these quality characteristics, and thus their functional relations, are imprecise to some degree due to tolerances, nonspecifi city, measuring errors, incomplete knowledge, vagueness of definitions , and so on. Here, possibility distributions and possibilistic regress ions are used to model these imprecise natures and induced imprecise f unctional relationships. We first integrate and extend existing possib ilistic regression methods to obtain unified measures of predictive qu ality characteristics or responses. The unified possibility distributi ons are also practical for many forecasting problems. We then propose a multiple objective programming model to obtain an appropriate combin ation of process parameter settings based on the unified possibility d istributions of imprecise predictive responses. We not only optimize t he most possible responses values, but also mini mite imprecision or d eviations from the most possible values. Through a die casting example , we show how to use our approach to reach an appropriate machine sett ing which simultaneously optimizes both porosity and temperature diffe rence on die surfaces.