統計製程管制（ SPC ）的主要功能之一是當製程有干擾項產生時，它能盡快的偵 測出來。 傳統之 SPC 管制圖在監督製程上有一重要假設：觀測值必須互相獨立，否則，當 自我相關性存在時，將造成假警訊之提高。 一般而言，針對 SPC 管制圖監督自我相關性資 料的處理方法為：(1) 利用自我迴歸整合移動平均模式描述自我相關性結構，而後建立殘差 值管制圖監督製程；(2) 配適指數加權移動平均模式，進而建立移動中心線指數加權移動平 均管制圖；(3) 整合工程製程管制，進而以 SPC 管制圖監督獨立之輸出偏離值。 本研究除 了推導理論說明上述三種方法之優缺點外，並配合模擬試驗，輔助說明之。

　　　　　The primary function of SPC charts is to detect the presence of disturbances as soon as when they introduce in the process. One important assumption for the traditional SPC charts requires that the monitored observations are independent to each other. Otherwise, the so called "false alarm" would be increased, and these improper signals result in the wrong interpretation and decrease the capability of SPC charts. The typical solutions for charting correlated observations can be described as the following three approaches. The first approach is in light of ARIMA modeling, and the residuals are charted by the traditional SPC charts. The second approach uses the MCEWMA charts to monitor process outputs. In the third approach, the suitable EPC is employed to tune the process to produce the independent output deviations from the target, and then the traditional SPC charts are used to monitor these independent observations. This study discusses the features of these three typica: approaches, and the limitations of these approaches are investigated. This study also addresses and evaluates the pros and cons of these three approaches by a series of simulations.