| Summary: | 碩士 === 國立交通大學 === 工業工程與管理系 === 88 === Department of Industrial Engineering and Management
National Chiao Tung University
Abstract
Process capability analysis means utilizing process capability indices to evaluate suppliers'' process capability when the processes are normally distributed and in control. It can be used for controlling and assessing process at the same time. Cp and Cpk are developed in early period and are usually used by industrial. Taguchi loss function concept is added into newer process capability index Cpm due to the lack of considering situation of process mean deviating from target value. Because the bigger the index doesn''t guarantee the higher the process yield, therefore, process capability index-Cpmk reflecting both the process yield and the expected loss is proposed.
Cpmk is rarely utilized so far. Though the table of 100(1-α)% confidence lower limit for Cpmk under different sample size and value when the specification interval is symmetry, the method is not practical owe to the assumption of large sample size. In addition, there is only hypothesis about the difference between two processes or suppliers'' process capability indices Cp.
Hence, the purpose of this study is using bootstrap simulation to construct confidence lower limit for single process capability index Cpmk and confidence interval for the difference between two process capability indices Cpmk. These can be utilized to find a better process or supplier by replacing traditional hypothesis. To facilitate and encourage the use of the engineer without statistic background, a program is provided for computing the confidence lower limit and confidence interval. Finally, a example is used to demonstrate the use of the proposed method.
|
|---|
