| Summary: | 碩士 === 逢甲大學 === 工業工程與系統管理學研究所 === 95 === Design of experiments and the Taguchi method are often applied to resolve single-response quality problems by quality engineers in many industries. For some industries such as shoe industry, IC package industry, and the like, multi-response quality problems are prevailing. However, the conventional design of experiments and Taguchi method are incapable of solving the multi-response quality problems. In this study, the data envelopment analysis method is proposed to resolve the multi-response quality problems. First, PCA method is applied to design combinations for multi-response quality problems. Next, variants of data envelopment analysis model are applied to provide solutions to the problems. Finally, results from data envelopment analysis methods and PCA method are compared and optimal solutions are identified. The proposed methodology is implemented using real-world data collected from a local IC assembly house. Results reveal that the proposed methodology outperforms the PCA method.
|
|---|
