跳到主要內容

臺灣博碩士論文加值系統

(44.220.44.148) 您好!臺灣時間:2024/06/14 12:11
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:林正宏
研究生(外文):CHENG-HUNG LIN
論文名稱:多變量統計用於製紙流程的錯誤診斷與分析
論文名稱(外文):Fault detection and classification based on multi-variant statistics in papermaking process
指導教授:陳炳宏陳炳宏引用關係
指導教授(外文):BING-HUNG CHEN
學位類別:碩士
校院名稱:國立東華大學
系所名稱:電機工程學系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2006
畢業學年度:94
語文別:中文
論文頁數:74
中文關鍵詞:錯誤診斷多變量主成分分析
外文關鍵詞:Principal Component Analysis
相關次數:
  • 被引用被引用:1
  • 點閱點閱:274
  • 評分評分:
  • 下載下載:59
  • 收藏至我的研究室書目清單書目收藏:1
現代化的產品及其生產過程日趨複雜,一個產品或製程的品質往往取決於兩個或兩個以上彼此相關的品質特性,所以利用單變量管制圖做個別的監控容易產生誤判的機會,故採用多變量管制圖來同時監控數個產品品質特性就顯得相當需要而且重要了。
主成分分析法屬於多變量投影法利用統計與線性代數的原理,將諸多可能具相關性變數空間,投影至較低維度且完全線性獨立的潛存變數空間,以擷取大量變數產生龐大數據隱含的特徵訊息。且透過模型的建立與驗證,對製程正常與異常操作狀態作明確的界定。本論文主要利用多變量統計的主成份分析,針對製程所產生的變數,進行主成份模型的建立,再分別建構錯誤診斷系統,即時掌握整體製程變數的變異狀態,並加以驗證。
本架構是以資料矩陣大小為50 (量測數目) × 12 (變數數目),建立主成份模型,再針對變數變化建立一個錯誤診斷系統,此系統可偵測之靈敏度範圍為標準值的 ± 3 %,將其分為二部份來驗證錯誤診斷,第一部份取一個變數並設定其標準值上限及下限的3 %,其結果在單一變數單一錯誤情形下,D²分佈圖和T²分佈圖的靈敏度都超越上限值的1.52及10.5,但在單ㄧ變數同時兩個錯誤發生時,T²分佈圖的靈敏度不如D²分佈圖;第二部份同時取二個變數並設定其標準值上限及下限的3 %,其結果在兩個變數單一錯誤情形下,D²分佈圖和T²分佈圖的靈敏度都超越上限值的1.52及10.5,而兩個變數同時兩個以上錯誤,其D²分佈圖及T²分佈圖都很明確顯示其靈敏度,因此在多個變數情形下,越可明確的反映其靈敏度,準確的診斷出錯誤與實際變化呈現相符合之現象。
It could be a very complicated process of manufacturing industrial products, their quality is usually decided by two or more related characteristics of process. Today, most production lines use a single variable diagram method to monitor the process parameter variation to correlate their quality, but have limited yield improvement. It is therefore necessary and important to develop a multi-variables diagram on monitoring several characteristics of product. The purpose of this work is to build a model can detect fault with sensitivity 3%.
Principal Component Analysis (PCA) is one of Multivariate Projection Methods, using principles of statistics and linear algebra to project variable spaces of samples in possible to Latent spaces which have low dimensions and linear independent. We set variables and transfer to eigenvector space, which corresponding eigenvalues mean statistic weighting, these space configured by so-call principle components. Through building and verifying models in raw data, it could determine fault process occurrence. This work analyses these variables of papermaking process and builds fault detecting model by using process parameters of 50 golden products which have good quality, then run multivariable statistics to monitor the real-time state of processes.
Model was build from 50 (counts of measure) × 12 (counts of variables) data matrix and successfully detect two kinds of faults, one with single variable within 3% variation, the other with two variables within 3% variation. In the future, we are going to develop the dynamic average limitation which automatically update model golden samples to overcome the process condition shift issues.
中文摘要.................................................Ⅱ
英文摘要.................................................Ⅲ
致謝.....................................................Ⅳ
目錄.....................................................Ⅴ
圖表目錄.................................................Ⅶ
第一章 序論
1.1 研究概要............................................1
1.2 文獻回顧............................................2
1.3 論文架構............................................4
第二章 多變量的主成份分析
2.1 主成份分析..........................................7
2.1.1 主成份分析之介紹................................7
2.1.2 主成份分析的主要性質...........................11
2.1.3 主成份分析的幾何解釋...........................11
2.2 主成份分析的監控圖表- T² 管制圖....................19
第三章 研究方法與架構
3.1 變數簡介及檢驗方法.................................23
3.2 資料分析工具.......................................40
第四章 數據分析
4.1 主成份模型.........................................42
4.2 錯誤診斷模型.......................................44
第五章 結論與未來發展
5.1 結論...............................................58
5.2 未來發展...........................................59
參考文獻 ................................................61
附錄一...................................................69
附錄二...................................................72
附錄三...................................................74
[1].Geladi, P., and Kowalski, B. R. “Partial Least-Squares Regression: A Tutorial,” Anal. Chim. Acta, vol. 185, pp. 1–17, 1986.
[2].Massy, W. F. “Principal Components Regression in Exploratory Statistical Research,” J. of the American statistical Association, vol. 60, pp. 235–256, 1965.
[3].Simoglou, A., and Martin, E. B. “Dynamic Multivariate Statistical Process Control Using Partial Least Squares and Canonical Variate Analysis,” Computers chem. Engng., no. Suppl., pp. S277–S280, 1999.
[4].Jackson, J. E. A User’s Guide to Principal Components, John Wiley & Sons, Inc., New York, 1991.
[5].Martin, E. B., and Morris, A. J. “Batch Process Monitoring for Consistent Production,” Computers chem. Engng., vol. 20, no. Suppl., pp. S599–S604, 1996.
[6].Nomikos, P., and MacGregor, J. F. “Monitoring Batch Processes Using Multiway Principal Component Analysis,” AIChE J., vol. 40, no. 8, pp. 1361–1375, 1994.
[7].Wise, B. M., and Gallagher, N. B. “The Process Chemometrics Approach to Process Monitoring and Fault Detection,” J. Proc. Cont., vol. 6, no. 6, pp. 329–348, 1995.
[8].MacGregor, J. F., and Jaeckle, C. “Process Monitoring and Diagnosis by Multiblock PLS Methods,” AIChE J., vol. 40, no. 5, pp. 826–838, 1994.
[9].Wachs, A., and Lewin, D. R. “Improved PCA Methods for Process Disturbance and Failure Identification,” AIChE J., vol. 45, no. 8, pp. 1688–1700, 1999.
[10].Ku,W., and Storer, R. H. “Disturbance Detection and Isolation by Dynamic Principal Component Analysis,” Chemometrics Intell. Lab. Syst., vol. 30, pp. 179–196, 1995.
[11].Wachs, A., and Lewin, D. R. “Improved PCA Methods for Process Disturbance and Failure Identification,” AIChE J., vol. 45, no. 8, pp. 1688–1700, 1999.
[12].Bakshi, B. R. “Multiscale PCA with Application to Multivariate Statistical Process Monitoring,” AIChE J., vol. 44, no. 7, pp. 1596–1610, 1998.
[13].Zhang, H., and Tangirala, A. K. “Dynamic Process Monitoring Using Multiscale PCA,” In Proceedings of the 1999 IEEE Canadian Conference on Electrical and Computer Engineering (Edmonton, Alberta, Canada, May 1999), pp. 1579–1584.
[14].張海潮, 2001 , “Lagrange 乘數法", 數學知識EPISTE MATH,http://episte.math.ntu.edu.tw/entries/en_lagrange_mul/index.html.
[15].N. D. Tracy et al., Multivariate control charts for individual observations, Journal of Quality Technology, 24, 88-95 (1992).
[16].Scott,W.E. and Abbott,J.C.,”Properties of Paper:An Introduction”,pp.89-110,TAPPIPress,1995
[17].郭蘭生,”色紙的顏色管理”,經濟部工業局八十九年度工業技術人才培訓計畫,pp.25-52,台灣區造紙工業同業公會,2000
[18].林師模, 陳宛欽,“多變量分析” 2003.
[19].Yang Yinghua, Lu Ningyun, Wang Fuli, Ma Liling, “A New FaultDetection and Diagnosis Method based on Principal Component Analysisin Multivariate Continuous Processes,” Intelligent Control andAutomation, 2002. Proceedings of the 4th World Congress on Volume4, 10-14 June 2002, pp. 3156–3160.
[20].Anna M. Ison, Costas J. Spanos, “Robust Fault Detection and FaultClassification of Semiconductor Manufacturing Equipment, ” Departmentof EECS, University of California, Berkeley.
[21].Hotelling, H., “Multivariate Quality Control,” Techniques of Statistical Analysis,Edited by Eisenhart, Hastay and Wallis, McGraw-Hill, New York, NY, pp.111-184 (1947).
[22].Shewhart,W. A. Economic Control ofQuality of Manufactured Product, Reinhold Company,Princeton, N. J., 1931.
[23].Page, E. S. “Continuous Inspection Schemes,” Biometrika, vol. 41, pp. 100–115, 1954.
[24].Srivastava, M. S. (1997). “CUSUM procedure for Monitoring Variability”.Communications in Statistics - Theory and Methods, 26(12), pp.2905-2926.
[25].Roberts, S.W. “Control Chart Test Based on Geometric Moving Averages,” Technometrics,vol. 1, pp. 239–250, 1959.
[26].Steiner, S. H. (1999). “EWMA Control Charts with Time-Varying Control Limits and Fast Initial Response”. Journal of Quality Technology, 31(1),pp. 75-86.
[27].H. Wold, Estimation of Principle Components and Related Models by Iterative Least Squares, Multivariate Analysis, R. P. Krishnaiah, ed., Academic Press. New York (1966), p.391.
[28].S. J. Qin and T. J. McAvoy, A data-based process modeling approach and its applications, in Proceedings of the 3rd IFAC Symposium on Dynamics and Control of Chemical Reactors, Distillation Columns, and Ba tch Processes. Pergamon, Oxford (1992a).
[29].J. Kresta, Application of Partial Least Squares Regression, PhD Thesis, McMaster Univ., Hamilton, Ont., Canata (1992).
[30]. B. M. Wise, Adapting Multivariate Analysis for Monitoring and Modeling Dynamic Systems, PhD Thesis, Univ. of Wishington, Seattle (1991).
[31]. S. J. Qin, Partial Least Squares Regression for Recursive System Identification, Proce. Conf. on Decision and Control, IEEE, Piscat-away, NJ (1993).
[32].P. Geladi and B. Kowalski, Partial Least-Squares Regression: A tutorial,Analytica Chimica Acta, 185 (1986) 1-17.
[33].R. Manne, Analysis of Two Partial Least Squares Algorithms for MultivariateCalibration, Chemometrics and Intelligent Laboratory Systems 2(1987) 283.
[34].A. Hoskuldsson && , PLS Regression Methods, Journal of Chemometrics 2(1988) 211.
[35].Inge. S. Helland, Some theoretical aspects of partial least squares regression,Chemometrics and Intelligent Laboratory Systems 58(2001) 97-107.
[36].B. M. Wise et al., A theorical basis for the use of principlal component models for monitoring Multivariate process, Process Control and Quality, (1990) 1,pp.41-51.
[37].W. Ku et al., Disturbance detection and isolation by dynamic principal component analysis, Chemometrics and Intelligent Laboratory Systems, 30 (1995)179-196.
[38].J. F. Macgregor and T. Kourti, Statistical process control of multivariate processes, Control Eng. Practice, vol. 3, No. 3, pp. 403-414 (1995).
[39].U. Kruger et. Al., An Alternative PLS Algorithm for the Monitoring of Industrial Process, Proceedings of the American Control Conference, Arlington, VA June 25-27, 2001.
[40].Anderson, T. W. (2003). An Introduction to Multivariate Statistical Analysis (3nd Edition), John Wiley & Sons, New York.
[41].Chen, L. H. and Wang, T. Y., “Artificial neural networks to classify mean shifts from multivariate x2 chart signals,” Computer & industrial engineering, 47,195-205 (2004).
[42].Huwang, L., Yeh, A. B., and Wu, C. (2005). “Monitoring Multivariate Process Variability for Individual Observations”. Manuscript, pp. 1-28.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top