本論文探討當生產設備壽命屬韋伯分配時,整合預防保養與統計製程管制用於設計指數加權移動平均管制圖。討論有預防保養,和無預防保養的指數加權移動平均管制圖是否有差異。本研究首先針對考量預防保養下,求得每單位時間下最少成本之對應參數,樣本大小(Nm)、第 時段含抽樣及檢驗(hi)、管制界限(k)、以及保養次數(m)。數值分析結果顯示當Nm=551、h1=148、k=2.615、m=4的情況下,最佳之預防保養目標值E(C)=60.1422。設備壽命時間屬於韋伯分配,韋伯分配參數值愈大者,會有愈大的目標值。另以模擬試產生製程數據評估ARL,結果顯示有執行預防保養的統計製程管制EWMA管制圖,比無執行預防保養的統計製程管制EWMA管制圖成本較高但ARL較長。
In this study, we integrated preventive maintenance and SPC to design a EWMA control chart. We assumed the life time of a manufacturing machine follows Weibull distribution. The main concern is to understand the difference of designing EWMA control charts with PM or without PM. The minimum cost per unit time based on the corresponding parameters, the sample size (Nm), the first time with the sampling and testing (hi), control limits (k), as well as maintenance frequency (m) value is cakulated for preventive maintenance. Numeral result shows that the best preventive maintenance target value E(C) = 60.1422 attains with Nm = 551, h1 = 148, k = 2.615, m = 4.Also, a higher parameter value of Weibull distribution is a larger target E(C) has. Though, a EWMA control chart with preventive maintenance consideration, costs higher than EWMA control chart without PM but the ARL0 is longer than the one without PM.