Bias-specified robust design optimization: A generalized mean squared error approach

Authors:
Gunther Steenackers;Patrick Guillaume
Affiliations:
Vrije Universiteit Brussel (VUB), Acoustics and Vibration Research Group (AVRG), Department of Mechanical Engineering (MECH), Pleinlaan 2, B-1050 Brussel, Belgium;Vrije Universiteit Brussel (VUB), Acoustics and Vibration Research Group (AVRG), Department of Mechanical Engineering (MECH), Pleinlaan 2, B-1050 Brussel, Belgium
Venue:
Computers and Industrial Engineering
Year:
2008

Citing 2
Cited 2

MATLAB 5 for Engineers

MATLAB 5 for Engineers
Bias-specified robust design optimizaion and its analytical solutions

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

Development of an adaptive response surface method for optimization of computation-intensive models

Computers and Industrial Engineering
Extending the desirability function to account for variability measures in univariate and multivariate response experiments

Computers and Industrial Engineering

Quantified Score

Hi-index	0.00

Visualization

Abstract

The primary goal of robust design is to determine the best design points by minimizing variability of the performance of a manufacturing process and minimizing deviation of the performance from a target value of interest. In this paper, a generalization and improvement of the robust optimization approach, based on the mean squared error, is suggested and applied. The constraint on the process bias is further relaxed. Different response surface functions are evaluated and compared. As an example, different optimization approaches are illustrated with a numerical test case. The results are explained and discussed.