A. Miller et Cfj. Wu, PARAMETER DESIGN FOR SIGNAL-RESPONSE SYSTEMS - A DIFFERENT LOOK AT TAGUCHIS DYNAMIC PARAMETER DESIGN, Statistical science, 11(2), 1996, pp. 122-136
A recent trend in the industrial applications of robust parameter desi
gn is to consider complex systems which are called ''systems with dyna
mic characteristics'' in Taguchi's terminology or signal-response syst
ems in this paper. This potentially important tool in quality engineer
ing lacks a solid basis on which to build a rigorous body of theory an
d methodology The purpose of this paper is to provide such a basis. We
classify signal-response systems into two broad types: measurement sy
stems and multiple target systems. Three issues are then of fundamenta
l importance. First, a proper performance measure needs to be chosen f
or system optimization, and this choice depends on the type of system.
Taguchi's dynamic signal-to-noise ratio is shown to be appropriate fo
r certain measurement systems but not for multiple target systems. Sec
ond, there are two strategies for modeling and analyzing data: perform
ance measure modeling and response function modeling. Finally, the pro
per design of such experiments should take into account the modeling a
nd analysis strategy. The proposed methodology is illustrated with a r
eal experiment on injection molding.