The objective of the Taguchi parameter design is to improve the performance of a product or process by determining the levels of its design parameters such that the performance characteristic is robust against various causes of variation. For this purpose, Taguchi proposed an experimental method in which a product array is used as an experimental design and a performance measure called the SN ratio is employed for analyzing the experimental data.
Parameter design problems are broadly classified into static and dynamic ones depending on whether a signal parameter exists or not. Static problems are further classified into the-nominal-the best (NB), the-larger-the-better (LB) and the-smaller-the-better (SB) ones, and for each problem type a unique performance measure (or an SN ratio) is appropriately defined. However, for the dynamic parameter design problem, such a classification has not been explicitly made. In this thesis, dynamic parameter design problems are also classified into DNB, DLB and DSB ones with the target slopes being $ 0 < \beta_{t} < \infty $, $ \infty $, and 0, respectively.
For dynamic parameter design problems, Taguchi proposed a two-step procedure which consists of a maximization step for an SN ratio and an adjustment step by an adjustment parameter. Although this two-step procedure is very useful in practical applications, the employment of an SN ratio as a performance measure in the first step is not always justified.
The parameter design for the static case has been extensively studied by many researchers, and various alternative approaches have been proposed. However, relatively little literature exists for the dynamic case in spite of its practical importance. In addition, since the existing approaches for dynamic parameter design problems have some limitations, a variety of dynamic parameter design problems cannot be effectively studied.
This thesis deals with parameter design methods for dynamic systems.
For an active ...