Monitoring of profiles with within-profile correlation

• Main Authors: Jin-Pin Lin, 林巾平
• Other Authors: Chuen-Sheng Cheng
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/15482492307734381571

碩士 === 元智大學 === 工業工程與管理學系 === 104 === A basic assumption in statistical process control (SPC) applications is that quality characteristics can be well represented by a single measurement from a univariate distribution or multiple measurements from a multivariate distribution. Control charts are a commonly used tool in SPC to monitor the quality characteristics over time. In some practical applications, the quality of a product or process can be characterized by a relationship between a response variable and one or more explanatory variables that is referred to as a profile. Most existing profile monitoring methods are based on the assumption that the observations within each profile are independent of each other. This assumption is often invalid in practice. Successive measurements within profiles often exhibit spatial or serial correlation. This research focuses on profile monitoring when within-profile data are correlated. We propose the use of time series analysis and wavelet analysis to account for correlation structure within a profile. Relevant feature vectors are proposed to construct the support vector machine based classifier. Simulation results indicated that the proposed method can improve the classification accuracy of SVM.