IDENTIFICATION OF QUALITY CONTROL VARIABLES IN GLASS MANUFACTURING

摘要

Glass manufacturing is a process in which there are a greatrnnumber of input variables. The effects of these variables are notrnfully understood. In addition, the process occurs over severalrnhours. Quality defects in glass are observed when the glassrncools, often as much as a day after the event that caused therndefect. Modern technology has brought to glass manufacturingrnthe ability to collect a great amount of data about the process.rnHowever, modeling the data by conventional methods has notrnbeen successful. This paper presents a new method of analysis.rnThe paper focuses on the method rather than glassrnmanufacturing and its particular engineering. The method isrngeneral and can be applied to any process where a large amountrnof data has been collected.