Computing trade-offs in robust design: Perspectives of the mean squared error

Authors:
Sangmun Shin;Funda Samanlioglu;Byung Rae Cho;Margaret M. Wiecek
Affiliations:
Department of Systems Management & Engineering, Inje University, Gimhae, KN 621-749, South Korea;Department of Industrial Engineering, Kadir Has University, Kadir Has Campus, Cibali 34083, Istanbul, Turkey;Department of Industrial Engineering, Clemson University, Clemson, SC 29634, USA;Department of Mathematical Sciences, Clemson University, Clemson, SC 29634, USA
Venue:
Computers and Industrial Engineering
Year:
2011

Citing 6
Cited 1

Performance measures independent of adjustment

Technometrics
Signal-to-noise ratios, performance criteria, and transformations

Technometrics
Taguchi's parameter design: a panel discussion

Technometrics
Augmented Lagrangian and Tchebycheff Approaches in Multiple Objective Programming

Journal of Global Optimization
Robust design modeling and optimization with unbalanced data

Computers and Industrial Engineering
Bias-specified robust design optimization and its analytical solutions

Computers and Industrial Engineering - Special issue: Selected papers from the 31st international conference on computers & industrial engineering

Comparative studies on the high-variability embedded robust parameter design from the perspective of estimators

Computers and Industrial Engineering

Quantified Score

Hi-index	0.00

Visualization

Abstract

Researchers often identify robust design as one of the most effective engineering design methods for continuous quality improvement. When more than one quality characteristic is considered, an important question is how to trade off robust design solutions. In this paper, we consider a bi-objective robust design problem for which Pareto solutions of two quality characteristics need to be obtained. In practical robust design applications, a second-order polynomial model is adequate to accommodate the curvature of process mean and variance functions, thus mean-squared robust design models, frequently used by many researchers, would contain fourth-order terms. Consequently, the associated Pareto frontier might be non-convex and supported and non-supported efficient solutions needs to be generated. So, the objective of this paper is to develop a lexicographic weighted-Tchebycheff based bi-objective robust design model to generate the associated Pareto frontier. Our numerical example clearly shows the advantages of this model over frequently used weighted-sums model.