An application of robust parameter design using an alternative to Taguchi methods

摘要

The factors of interest in designing a product or process can generally be classified into two categories, controllable and uncontrollable. Controllable (or control) factors represent those factors which can be regulated. Examples of control factors include: the choice of material, flow rates, processing pressures, times and temperatures. Uncontrollable (noise) factors are those that are either difficult, impossible or too expensive to control during actual production or use. Examples of noise factors are: environmental conditions such as ambient temperature or humidity, process parameters which are dictated by an outside source such as end user demand, and usage factors such as how long and at what temperature a consumer stores a product. As compared to the current Tagachi approach, a new design method which provides greater flexibility in the design of the experiment, utilize a more meaningful performance statistic, and lend itself to a better understanding of the product or process is described in this paper.