| Summary: | 碩士 === 國立勤益技術學院 === 工業工程與管理系 === 93 === Quality index, SPK, holding one-to-one mathematical relationship with the process yield, can not only evaluate the capability but also reflect the yield of a process. Because the distribution function of the estimator of index SPK is quite complex, it is difficult to infer it’s lower confidence limit. Since the errors induced from sampling, it is not objective to use only the point estimation of the index to evaluate a process quality. Because the index SPK is the function of unilateral specification indices CPU and CPL, this paper will derive the lower confidence limit of index SPK form the lower confidence limit of indices CPU and CPL, and utilize the Monte Carlo simulation approach to evaluate the coverage percentage in order to verify the accuracy of this new approach. With the lower confidence limit of index SPK, the manufacturing industry can use it to evaluate or test whether toe quality meets the preset demands.
|
|---|
