Robust Design Using Preference Aggregation Methods

Robust design is a methodology for improving the quality of a product or process by minimizing the effect of variations in the inputs without eliminating the causes of those variations. In robust design, the best design is obtained by solving a multicriteria optimization problem, trading off the nominal performance against the minimization of the variation of the performance measure. Because these methods often combine the two criteria with a weighted sum or another fixed aggregation strategy, which are known to miss Pareto points, they may fail to obtain a desired design. To overcome this inadequacy, a more comprehensive preference aggregation method is combined into robust design. Two examples are presented to illustrate the effectiveness of the proposed method.Copyright © 2003 by ASME