抄録
TM, Taguchi Methods, uses OA Orthogonal Arrays, and ANOVA, Analysis of Variance, as in DOE, Design of experiments. The difference between TM and DOE was discussed in this article. TM uses OA as an inspection device to check existance of interactions among control factors by assigning those factors to OA so that main effects are firmly confounded with various interactions as much as possible. Because TM is intcrested in the reproduciability of technology developed within laboratories by using mostly specimens or simulation models on actual products. The S/N ratio used in TM are based on the mathematical analysis called quadratic expansion,more generally Hermitian form expansion, but not on regression analysis used in traditional DOE. TM also strongly recomends to usc a dynamic S/N ratio to measure the functionality of generic function. TM docs not want to use a static S/N ratio related with objective function or any quality cahracteristics, as far as possible.
