透過您的圖書館登入
IP:3.144.42.196
  • 學位論文

應用類神經網路監控多變量製程之變化與參數估計

Monitoring and Characterizing the Changes of Mean and Variance in a Multivariate Process by Artificial Neural Networks

指導教授 : 鄭春生
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


在統計製程管制中,當製程受到多個具有相關性之品質特性所影響時,稱為多變量品質管制問題。由於科技之進步,先進的自動化檢測設備能同時量測一件產品中多項品質特性,若同時以多個單變量管制程序監控此多項品質特性,將由於特性間的相關性而造成過大的判斷誤差。過去有許多學者應用統計原理提出多種不同的多變量製程管制程序,以偵測多變量製程平均值或變異性之變異情況,但其效益仍有改善空間。因此,多變量製程管制程序的發展與其效益之提升,已是刻不容緩。 本研究以類神經網路為基礎,發展一監控多變量製程平均值與變異性變化之系統,此系統並可進一步估計平均值偏移量的大小。本研究採用不同的網路輸入向量供類神經網路進行學習。在製程偏移之偵測與偏移量之估計上,以平均連串長度與絕對誤差百分比來評估類神經網路的偵測能力,並與傳統管制法進行比較。此外,本研究並以模擬數值範例驗證類神經網路管制法之成效。研究成果顯示,類神經網路的偵測效益優於傳統管制法。 本研究的主要貢獻有三點。第一、本研究發展之類神經網路管制法的效益,並不會因共變異矩陣 ( ) 不同而有所差異,只要多個變量之間所產生之偏移量 ( ) 相同,則具有相同的偵測效益。第二、傳統的多變量CUSUM管制法有極佳之偵測效益,但其必須在偏移量已知的情況下使用,才能達到最佳的偵測效益。而本研究所提出之類神經網路管制法,可監控較大範圍之製程偏移量,使其在製程偏移量未知之情況下,偵測效益優於傳統之多變量CUSUM管制法。第三、類神經網路管制法對於製程偏移量的估計優於傳統管制法。製程偏移量之估計,將有助於工程人員擬定製程調整之對策。

並列摘要


Quality-control problems in several related variables are called multivariate quality-control problems. Various types of quality control charts have been proposed to monitor the mean or variance of a multivariate process. Besides the use of control charts in monitoring process and identifying assignable causes, quality practitioners frequently need to adjust processes based on the magnitude of change. In this research, we develop a control procedure based on artificial neural network to monitor the mean and variance of the multivariate process. At the same time, this procedure will predict the magnitudes of changes to provide the engineers to modify the process. The performance of the artificial neural network and traditional multivariate quality control method have been evaluated by the average run length (ARL) and mean absolute percent errors (MAPE) using simulation. The results indicate that artificial neural network has certain advantage over the traditional multivariate quality control method. The feature of this research is that even if the magnitude of changes is unknown, the detecting ability of neural network is better than traditional multivariate CUSUM method.

參考文獻


3. Anderson, T. W., “An introduction to multivariate statistical analysis,” John Wiley & Sons, New York (1958).
4. Aparisi, F., Jabaloyes, J. and Carrion, A., “Statistical properties of the
5. Chang, S. I. and Aw, C. A., “A neural fuzzy control chart for detecting and classifying process means shifts,” International Journal of Production Research, 34, 2265-2278 (1996).
6. Chen, L. H. and Wang, T. Y., “Artificial neural networks to classify mean shifts from multivariate chart signals,” Computer & industrial engineering, 47, 195-205 (2004).
7. Cheng, C. S., “A multi-layer neural network model for detecting changes in the process mean,” Computers and Industrial Engineering, 28, 51-61 (1995).

被引用紀錄


阮冰如(2006)。應用類神經網路與支援向量機於多變量製程 變異來源之辨識〔碩士論文,元智大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0009-1807200617004500
黃威榮(2009)。應用類神經網路於高產出製程監控之研究〔碩士論文,元智大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0009-3007200909503300

延伸閱讀