Supersaturated split-plot designs for robust parameter experiments

Abstract

The intention of robust parameter design (RPD) experiments is to identify the influence of control factors on the variation introduced by the uncontrollable noise factors. Split-plot designs can be effectively applied to decrease the cost in RPD experimentation when there are hard-to-change or hard-to-control factors. In the presence of many factors, either the traditional replicated split-plot designs or the Cartesian-product designs proposed by Taguchi are not of practical use since they require a large number of experimental units; thus, fractional factorial designs, such as supersaturated designs (SSDs) are needed. In this article, we propose a construction for supersaturated split-plot designs suitable for RPD experiments. We implemented the methodology of compound orthogonal arrays using orthogonal arrays of strength greater than 1, for the split-plot units and any relevant balanced SSD for the whole-plot units. The proposed methodology indicates that the experiments conducted in a split-plot point of view can be of tremendous value in RPD since they enable the estimation of the contrasts of interest in a considerably easier way than the designs proposed by Taguchi.

Keywords:

• compound orthogonal arrays
• SSDs
• robust parameter designs
• foldover design

About the authors

Dr. Drosou is a Doctor in the Department of Mathematics at National Technical University of Athens. Her email address is [email protected].

Dr. Koukouvinos is a Professor in the Department of Mathematics at National Technical University of Athens. His email is [email protected].