本文使用Nelder-Mead 的求解方法，求出累積值管制圖的各項變數之最經濟組 合，以替代Shewhart 管制圖，以期運用於實務中達到控制、管理製程的目的。 本研究工作發現： 1.韋氏分佈之尺度參數，對平均成本與平均週期時間的影響較形狀參數為大。 2.若欲偵測之製程偏離幅度增大，則應採用較小的樣本組，並增加抽樣次數。 3.運算間距在平均成本與PC運算時間上，對決策區域的影響較大；成本函數運算初 期，以較大運算間距較適宜，運算最後階段，則以較小運算間較適當。

　　　　　An economic model is developed which determines the design of Cumulative sum control chart to minimize costs. It provides an operational tool which will permit the Cumulative sum control chart to be used in an economically optimum manner as an alternative to Shewhart control charts for monitoring a process in a realistic environment. Three conclusions have been drawn:(i) the Weibull scale pa- rameter affects more the variation in loss-cost and cycle time than does the Weibull shape parameter; (げ ) small subgroup sizes should be taken more often when the magnitude of shift in the process mean, which is to be detected, increas- es; (こ ) the step size affects more the variation in decision interval than does the subgroup size and the sampling interval; it is suitable for the loss-cost function to take a large step size first and a small step size later.