| Summary: | 碩士 === 國立高雄應用科技大學 === 工業工程與管理系 === 99 === Traditionally, variable control charts are constructed based on precise data collected from well-defined quality characteristics of manufacturing products. However, in the real world there are many occasions the stated quality characteristic of products such as surface roughness of optical lens or light transmittance of touch screens contains somewhat degree of imprecise information so that their readings inevitably gather as fuzzy numbers (data). In this situation, the traditional method for constructing the variable control charts exists a limitation of dealing with fuzzy data when the average and variance of underlying quality characteristic are monitored. Therefore, in this paper we propose and control charts with fuzzy data, whose fuzzy control limits are obtained on the basis of employing two well-known principles, so called resolution identity and extension principle, in the fuzzy theory. Furthermore, for comparing fuzzy observations and fuzzy control limits, a fuzzy ranking method is presented to classify the underlying manufacturing process condition. Finally, a practical example is provided to demonstrate the applicability of our proposed methodologies.
|
|---|
