Determining the Contributors to SPC Charts Signal in a Multivariate Process

• Main Authors: Hsu Bo-Sheng, 許博昇
• Other Authors: Yuehjen E. Shao
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/25234251199955103768

碩士 === 輔仁大學 === 應用統計學研究所 === 96 === Quality is the most important feature to describe the products. In order to maintain the products’ quality, the quality control techniques are essential for process monitoring. Because the statistical process control (SPC) chart has practical monitoring capability, it becomes one of the most useful quality control techniques. The SPC charts are able to effectively and correctly detect the process disturbances when they were introduced in the process. Nevertheless, the SPC charts still have some limitations, especially in monitoring a multivariate process. A multivariate process has two or more variables (or quality characteristics) to be monitored. When an out-of-control signal is triggered by the SPC chart, the process personnel usually only know that the process is something wrong. Yet, it is hardly to determine which of the monitored quality characteristics is responsible for this out-of-control signal. In this study, we propose the machine learning mechanisms to solve this difficulty. We integrate the neural network (NN) and support vector machine (SVM) with the multivariate SPC charts to identify the contributors to an SPC signal. The fruitful results are demonstrated through the use of a series of simulations.