陳彥匡 博士邱紹瑜2025-08-112025-08-112006-08-31U0061-1811200915323653https://www.airitilibrary.com/Publication/alDetailedMesh1?DocID=U0061-1811200915323653由於適應性<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖比起傳統的<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖偵查非機遇原因速率較快，因而近年來已有許多學者針對適應性<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖的設計參數如何調整之議題深入作探討。設計參數包含有：樣本大小、抽樣間隔、管制界限或警告界限。為了使適應性管制圖的運作成本得以降低，有學者提出採經濟設計原則進行管制圖設計參數的設定，並且與傳統<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖的運作成本比較。雖然現有文獻的比較結果顯示，適應性管制圖的運作成本低於傳統管制圖，但其比較於法上，由於每次僅採更改一個製程或成本參數值並觀察兩者運作成本的變化；此種比較方式易於忽略參數間之交互作用效果，而可能產生不正確的比較分析結果。有鑑於此，本研究結合實驗設計方法與共變數分析的技巧，提出一採用兩階段分析程序，同時比較VP、VSSI、VSS、VSI等四種適應性<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖與傳統<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖。依此，本研究發現，選用不同類型的適應性或傳統<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖，其最低成本的高低並無顯著差異，因此以往學者認為「以適應性<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖取代傳統<img src="/theabs/image?id=001&account=095NTTI0163006">管制圖將可節省運作成本」的想法，可能須加以修正。Adapting the design parameters of the <img src="/theabs/image?id=001&account=095NTTI0163006"> control chart has been investigated extensively in recently years because it can cut down the time of detecting a small or moderate process shift as compared with the traditional <img src="/theabs/image?id=001&account=095NTTI0163006"> control charts. These design parameters include the sample size, the length of sampling interval, and the control/warning limits of a control chart. In order to operate the adaptive chart economically, several studies have devoted to the economical design of the adaptive chart. Numerical examples in these studies have shown that economically designed adaptive chart could be more cost-effective than the traditional one. However, these examples might be insufficient to convince the user because the results lack a rigidly statistical examination of this statement. As a result, this paper presents a two-phase analytic procedure based on the experiment design, the regression analysis, and covariance analysis to examine the difference among the traditional and some adaptive charts with one, two, or the whole of design parameters changeable in their cost performance. The analytic result indicates that the cost difference among these charts is minor.zh經濟設計、適應性管制圖、共變數分析、基因演算法、實驗設計economic design, adaptive control charts, analysis of covariance, genetic algorithms, factorial experiment design.master thesis