查詢結果分析
來源資料
頁籤選單縮合
題 名 | 監控非機遇性變異之自我相關製程之研究=Monitoring and Diagnosing the Autocorrelated Process by Transient Special-Caues Variations Using Covariance Components Analysis |
---|---|
作 者 | 陳慶文; 劉天賜; | 書刊名 | 品質學報 |
卷 期 | 12:1 民94.03 |
頁 次 | 頁79-87 |
分類號 | 494.56 |
關鍵詞 | 自我相關; 共變數成份分析; 隨機效果; Autocorrelation; Covariance components analysis; Random effect; |
語 文 | 中文(Chinese) |
中文摘要 | 在傳統統計製程品管,製程品質特性經常是相互獨立。然而某些非隨機性因素,例如操作人員的疲勞或機具長時期的磨損,會使製程產生自我相關結構之非機遇性變異,使得前後產品的品質特性產生關聯。等到該非隨機變異因素給排除後,製程品質特性就又會回到了原來相互獨立的狀態。此等製程之自我相關性,比較不適合使用時間序列分析。本研究以共變數成份分析為基礎,結合即時品管的迴溯性分析,以偵測製程因自我相關所引起的非機遇性變異,並加以排除。 |
英文摘要 | In traditional Shewhart control charts, the quality characteristics of a process is usually assumed to be independent. However, certain transient special-cause variations, such as workforce’s fatigue, would cause the quality characteristics in the process to be correlated resulting unexpected changes in the process. When the factors of transient special-cause variations are eliminated, the process will return to its original and independent production status. The typical time-series model is not appropriated for a process with autocorrelation by transient special-cause variations, which may shift the process’ mean for a short while, and may come back later. Incorporating the covariance components analysis, the study formalizes the quality control process to monitor and eliminate the autocorrelation by transient special-cause variations. |
本系統中英文摘要資訊取自各篇刊載內容。