Fuzzy Test for Taguchi Capability Index with Asymmetric Tolerances  
Author Kuen-Suan Chen


Co-Author(s) Chun-Min Yu


Abstract Taguchi capability index proposed by Chan et al. (1988) takes the process targeting issue into consideration, it fails to account for process with asymmetric tolerances. However, many tolerances are asymmetric in real-world practice to satisfy customer demand. Thus, Chen et al. (1999) proposed the Taguchi capability index with asymmetric tolerances to deal with processes that impose asymmetric tolerances. Unknown parameters must be formulated using estimates based on sample data, whereupon mathematical programming is used to determine the upper confidence limit of the resulting capability index. Unfortunately, cost considerations often limit the size of the sample that can be obtained and introduce measurement-related uncertainties, thereby compromising measurement results. In this study, we sought to enhance testing accuracy by applying an upper confidence limit of the capability index to derive a fuzzy number and membership function. We then developed a onetailed fuzzy test method to determine whether process capability reaches the required quality level. In numerical simulations, the proposed fuzzy testing method based on the upper confidence limit of the capability index was shown to enhance testing accuracy and thereby reduce the likelihood of misjudgment caused by sampling errors.


Keywords Taguchi capability index, asymmetric tolerances, fuzzy test, confidence interval
    Article #:  RQD28-66

Proceedings of 28th ISSAT International Conference on Reliability & Quality in Design
August 3-5, 2023