反應六標準差水準之多品質特性製程最佳化

Multi-Response Optimization for Achieving Six Sigma Level

王春和(Chung-Ho Wang)

《品質學報》 13卷3期 (2006/09) Pp. 289-304

摘要

產品品質為影響企業競爭力的重要因素，自從摩托羅拉(Motorola)公司成功以六標準差管理哲學來改善企業體質並獲取豐碩的利潤後，目前工業界皆以六標準差的品質水準為提昇產品品質的標竿。線外品質管制的技術是六標準差管理中改善產品品質極為有效的技術，其中，田口方法將雜音因子納入實驗規劃中，評估雜音因子對品質變異的影響，以找出具穩健性的最佳因子水準組合，其成效受到學術界的肯定並廣泛的應用於工業界。本研究將雜音因子納入直交表設計實驗中，從而發展出一套能反應六標準差水準的製程能力最佳化模式，以決定最佳的因子水準值。此最佳化模式利用多變量方法中的主成份分析法(PCA, Principal Component Analysis)處理多個品質特性具高度相關性的問題，並以反應模式(Response Model)估計雜音因子效應對品質特性的影響，最後利用六標準差望想函數(Desirability Function)，來評估多品質特性製程達六標準差水準之程度，從而決定最佳的因子水準值。本研究以密封元件油封(Oil-Seal)製程，來驗證所發展之反應六標準差最佳化模式之有效性。

關鍵字

直交表 ； 六標準差 ； 主成份分析 ； 望想函數

並列摘要

Product quality is a crucial factor for improving enterprise competitiveness in marketplace. Six Sigma has been widely used in industry to upgrade product quality at lower cost since Motorola company successful implementation of Six Sigma and create enormous profit. Off-Line quality improvement techniques such as design of experiments and Taguchi method is an important tool for improving product quality in the implementation of Six Sigma. Particularly, Taguchi Method has been extensively used in industry and obtains a promising result in improving product quality. Taguchi method accounts for the effects of noise factors on quality variation to establish the orthogonal array experiment and thereby determine an optimal factor level which is insensitive to the intervention of noise factors. Therefore, this study develops an optimization procedure to enhance product quality for achieving six sigma quality level based on Taguchi's orthogonal design. The principal component analysis from multivariate analysis is employed to resolve the correlation problem among multiple responses as constructing response model. The Six Sigma desirability function based on process yield is then employed to determine the overall quality performance for multiple responses and thereby determine the optimal parameters’ value to achieve the six-sigma quality level. An experiment regarding to the development of oil-seal demonstrates the effectiveness of the proposed procedure.

並列關鍵字

orthogonal table ； six sigma ； principal component analysis ； desirability function

參考文獻

Antony J.(2000).Multi-response Optimization in Industrial Experiments Using Taguchi`s Quality Loss Function and Principal Component Analysis.Quality and Reliability Engineering International.16,3-8.

Ching, C. K.,Quah, S. H.,Low, H. C.(2005).Index Cpm in Multiple Response Optimization.Quality Engineering.17,165-171.

Derringer, G.,Suich, R.(1980).Simultaneous Optimization of Several Response Variables.Journal of Quality Technology.12(4),214-219.

Hsieh, K.-L.,Tong, L.-I.(2001).Optimization of Multiple Quality Responses Involving Qualitative and Quantitative Characteristics by Applying Artificial Neural Networks.Computers in Industry.46(1),1-12.

Hsu, C. M.,Su, C. T.(2004).Simultaneous Optimisation of the Broadband Tap Coupler Performance Based on Neural Networks and Exponential Desirability Function.International Journal of Advanced Manufacturing Technology.23,896-902.