| Summary: | 碩士 === 逢甲大學 === 工業工程與系統管理學研究所 === 97 === The issue of selecting suppliers or assessing the impact of process improvement has become an extremely important role for manufacturers to increase their competitiveness and it has already been investigated based on several process capability indices. The third-generation index Cpmk index was constructed by appropriately combining the yield-based index Cpk and the loss-based index Cpm, accounting for the process yield as well as the quality loss. Nevertheless, the investigation of this problem based on index Cpmk has not been developed due to the complexity of the sampling distribution. In this thesis, we consider seven methods of constructing approximate confidence intervals for the ratio and difference between two Cpmk values. The seven methods are Bonferroni, Modified Bonferroni, Difference, Generalized Confidence Intervals (GCI) and three based on the Bootstrap, referred to as Standard Bootstrap (SB), Percentile Bootstrap (PB) and the Biased-Corrected Percentile Bootstrap (BCPB). Furthermore, the performance of these methods for selecting and judging the better of two suppliers or processes are investigated and compared via simulation. The results indicate that the Difference, GCI, BCPB on the ratio and SB on the difference provide better performance in most of our studied cases with stable error probability and better selection power. Finally, an application example is presented for decision makers as a reference to assess the capability of suppliers or processes.
|
|---|
