SPC與EPC整合系統在多變量自我相關製程之應用

Abstract

管制圖是統計製程管制（Statistical Process Control, SPC）中最常被工業界用來監控製程或產品品質之工具，可有效地偵測出影響製程的非機遇原因。但是當製程資料具有顯著的自我相關（autocorrelated)時，使用傳統的管制圖將會產生錯誤的訊息，造成企業成本的損失。然而，製程除了受到可歸屬原因的影響外，也可能受到不易控制的機遇原因（chance cause）所影響。當製程發生失控的情況且無法查出非機遇原因時，製程工程師大多使用工程製程管制（engineering process control, EPC）來進行回饋控制，藉由調整製程之輸入變數，使製程輸出值接近目標值。近年來，整合SPC與EPC的作法深受產業界之重視，透過整合SPC與EPC，可使製程管制的效果更臻理想。此外，由於現今產品功能複雜，多個品質特性之製程管制工作日受重視，但中、外文獻罕見有關多變量自我相關製程管制之相關研究。因此本研究針對多變量自我相關之製程，提出一套完整且有效之結合SPC與EPC之管制流程。本研究所提出之管制流程分為兩階段，第一階段為SPC程序，首先使用Z管制圖偵測製程平均數是否發生偏移。當製程失控卻找不到可歸屬原因時，則進行第二階段之EPC程序，利用自組性演算法(Group Method of Data Handling, GMDH)得到品質特性與控制變數之間的關係式，再透過此關係式得到各個控制變數的製程增益值（process gain)，並根據製程發生失控的品質特性之實際值與目標值之間的偏移量進行回饋控制，使製程之輸出能夠接近目標值。應用本研究所建立之管制流程可以提供製程工程師一套有效且準確之具自我相關特性多變量製程管制程序。本研究最後利用一個化學製程實例及模擬案例証實了本研究方法確實有效可行。

The control chart is a widely employed statistical process control (SPC) tool for monitoring the process. It can detect the assignable cause effectively. However, if the process has significant autocorrelation, the traditional SPC procedure would cause suspious information. Besides, chance causes also have impact on the processes. When the process is out of control but no assignable cause is found, it can be adjusted by employing engineering process control (EPC). Recently, the integrated SPC and EPC scheme is the gaining interests in industries. Superior control can be achieved through the application of integrated SPC and EPC scheme. This study presents an integrated SPC and EPC procedure for multivariate autocorrelated process. The SPC procedure constructs a control chart called Z-chart to monitor the process. In the EPC procedure, the modeling way of the group method of data handling (GMDH) is employed to construct the EPC model to adjust the process mean to the target value. GMDH is used to predict the characteristic values of a process output and the relationship between input and output, respectively. A case is utilized to demonstrate the effectiveness of the proposed procedure.