A Predictive Control Chart Based on Dynamic Bayesian Approach

• Main Authors: Keng, Wei, 耿威
• Other Authors: Wu, Ji-Cheng
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/41802216950367885984

碩士 === 國立屏東科技大學 === 工業管理系所 === 98 === The purpose of this paper is to study the making of dynamic Bayesian model and the application of the statistical process control using this dynamic Bayesian structure to monitor the process. In the history of the development of control charts, only few attempts have so far been made at Bayesian control charts. In real life situation, a good quality engineer can use his/her experience of the product process and the information of the sample to make a professional judgment. The state equation of dynamic Bayesian model is to describe the change of the process state parameter in different stage. In addition, using the observe equation is to reflect the relationship of the samples gathered from different stages of process state parameter. In this study, we propose a set of guidelines for the decision maker to dynamically monitor the changes in every stage of the process based a constant mean model of the dynamic Bayesian control chart. Finally, apart from explaining the theoretical structure, the average run length (ARL) will be used to evaluate the monitoring power of the dynamic Bayesian control chart. It is hoped that this study will not only be a basis for future academic research but will also offer practical help in effectively monitoring the process method, thus providing quality control engineer a timely and effective tool in evaluating process. In this way, it may help to produce high quality goods, enabling the company to be more competitive.