Accuracy Analysis of Measurement System

• Main Authors: Hsieh-chin Li, 李協親
• Other Authors: Wen-kuei Chen
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/64306025866763337363

碩士 === 義守大學 === 工業工程與管理學系碩士班 === 97 === In order to sustainable survival and development, and for quality management system requirements, modern enterprises need to develop and implement discipline, to ensure that the product quality. Recently to participate in the ISO 9000 standard certification, or the TS 16949 standard certification, and import the Six Sigma system, the enterprise has been pursuit the international quality criterion of supremacy. As a result of the measurement management is a key element of quality assurance system, modern enterprise should be the measurement system as the focus of inspection item. In1994, the Big Three U.S. automakers first jointly promulgated the QS 9000 standard, formally the Measurement System Analysis as a mandatory item. For in accordance to the ISO 9000:2000 standard, the Big Three U.S. automakers require the quality management system of suppliers will be fully converted to the TS 16949 standard since 2006. However, the MSA is still the requirements item of the system mandatory. As the North American Automotive Industry jointly request the use of MSA has more than 10 years, lead the industry general will be the MSA of the TS 16949 standard as the requirements item of quality management system. The MSA is based on the stability of system performance, and the main content of system performance contains precision and accuracy. The precision evaluation on the MSA, the method of various industries to adopt are similar; but，the accuracy evaluation method of adopted by various industries, it is no different, and many are troubled fallacy. This study investigated the accuracy evaluation method of various industries, including the method of the TS 16949 standard, MINITAB Package and the U.S. Intel Corporation, and exposes their fallacy. Starting from the bivariate normal distribution, this study to build the statistical model of MSA; then, according to the model can master the variation relations between parts value, measured value and appraiser bias. Final, for the accuracy evaluation fallacy of various industries, this study advance the improved methods of meet the needs of various industries; And, this study development the software model of accuracy evaluation, to support the business users.