Parameter Design Analysis And Optimization Of Mode Of Study

Genichi Taguchi's robust parameter design is a kind of design method of product quality in industry, this paper analyses parameter design's disadvantages, as well as the corresponding improvement technologies and alternative methods, so as to make the parameter design's researchers and practicians to hold the basic principles and realization methods of parameter design more properly and completely, and to exert parameter design's effects in reducing variations in products and processes. The main contributions of this paper are described as follow:1. From two aspects of Taguchi's quality concepts and parameter design's basic contents , parameter design theory have been analysed.2. Beginning with the shortcomings of S/N ratios method, other three techniques of PerMIAs and data transformation and response surface methodology which are alternatives to S/N ratios method have been analyzed. In addition, for an example of a thermal expansion instrument's robust design, it has been proved that surface methodology is an improvement to S/N ratios method.3. Through pointing out the limitations of Taguchi outer/inner orthogonal array experiment method, it has been studied that three methods: comprehensive noise factors ,split-plot experiment and direct modeling of response using combined arrays , are alternatives to the orthogonal array design. Then from three general sides, the comparations of orthogonal experiment and classical statistical experiment design have been made.4. From the economical angle, it has been proposed that the dual response surface methodology has more advantages than parameter design. And then the optimization strategies of parameter design and surface methodology have been analysed from two respects of minimizing process bias and minimizing process variation, simultaneously other four kinds of optimization strategies resolving the compromise between these two objectives have been studied. Finally it has been proved in the eye of an example: according to the decision-makers' differ preferences and the diverse design environments, one kind of optimization strategy may be selected to achieve the economical parameter design.