Investigations on robust design from the perspective of optimization

The issue of robust design optimization compromising statistical analyses and optimization techniques for the development of a production process has received much attention from researchers and practitioners for many years. Response surface methodology is accepted as a standard alternative to modeling imperfectly understood process relationships by estimating the response functions for process parameters. In particular, the dual-response approach and the mean squared error model are well-recognized robust design models. In this paper, we provide comprehensive studies associated with statistical analyses and optimization techniques for robust design optimization using four different robust design models, including those two models. A numerical example is conducted, and comparative studies are performed.