一階自我相關多變量簡單線性輪廓的監控

近年來，輪廓監控常用來監控產品品質的統計製程管制方法，當一產品的品質好壞可以用解釋變數和反應變數之間的函數關係來描述時，透過監控品質函數關係是否發生改變來判斷產品或製程品質是否穩定，此函數資料即稱為輪廓，而此種監控方法即稱為輪廓監控。在相關的文獻探討過程中，發現多數的文獻中模型都是假設隨機誤差項之間是獨立的，但是在實際的情況下，連續製程會造成輪廓間或輪廓內的相關性，且單變量管制圖常無法滿足現實生活的應用，因此本文針對多變量簡單線性模型且隨機誤差項之間存在一階自我相關性下，提出兩種新管制方法來監控此輪廓資料監控的效率，並與Soleimani 和Noorossana(2014) 所提出的的三種監控方法做比較。從模擬結果可以得知，我們所提出的管制方法監控效果比舊有的管制方法來的好。

關鍵字

多變量簡單線性輪廓；一階自我相關； MEWMA 管制圖；輪廓監控

並列摘要

In recent years, profile monitoring, a method of statistical process control, is often used to monitor the quality of a process or product. When the quality of the product can be described by a functional relationship between explanatory variables and response variables, one can determine whether the quality of a product or process is in-control by monitoring the stability of the functional relationship. The functional data is called a profile, and this kind of monitoring method is called profile monitoring. In the literature, most of the used models are assumed to be independent for the random errors. However, in real applications, the continuous process are often cause between or within correlation, and single variable control charts are often unable to meet the real applications, Therefore, in this article, we propose two control charting schemes to monitoring first-order autocorrelation multivariate simple linear profile data and compare with the schemes proposed by Soleimani and Noorossana(2014). By the simulation results, our proposed schemes have better performance than the existing schemes.

並列關鍵字

Multivariate simple linear profiles ； First order autocorrelation ； MEWMA control chart ； Profile monitoring

參考文獻

Google Scholar

Amiri, A., Sogandi, F., and Ayoubi, M. (2016).Simultaneous monitoring of correlated multivariate linear and GLM regression profiles in Phase II. Quality Technology & Quantitative Management, DOI:10.1080/03610918.2018.1429619.