The Decomposition of Signal-To--Noise Ratios with Application to the Robust Design

In this paper, it is discussed the decomposition of signal-to--noise ratios with application to the robust design. The conception of PerMIA, which gives the connection between the G.Taguchi's signal-to-noise ratios and his social average quadratic loss, is quoted. As an example, the problem in the CMCM-97 is analyzed using the approach and it can be concluded that this approach is more convenient in optimization procedure to parameter design.