Effective Pattern Recognition of Control Charts Using a Dynamically Trained Learning Vector Quantization Network

應用動態訓練提昇LVQ網路即時辨識管制圖異常形狀之績效

10.29977/JCIIE.200801.0008

顧瑞祥(Ruey-Shiang Guh)；薛友仁(Yeou-Ren Shiue)

靜態類神經網路 ； 形狀辨識 ； 管制圖 ； 統計製程管制 ； 學習向量量化網路 ； static artificial neural network ； pattern recognition ； control chart ； statistical process control ； learning vector quantization

工業工程學刊

25卷1期（2008 / 01 / 01）

73 - 89

英文

特定的管制圖異常形狀通常與特定的製程異常原因有關，因此，即時有效地偵測及分辨管制圖中出現的異常形狀，可幫助品管人員快速找到造成製程異常的原因，從而降低不良品發生的機率。近年來，有很多研究利用類神經網路線上即時辨識管制圖的異常形狀，然而，如何在即時辨識的條件下，精確地分辨異常形狀的形狀類別，是這一類研究常有的共同問題。這個問題來自於此領域絕大部份的研究皆採用靜態(static)監督式類神經網路作為辨識管制圖的工具，如倒傳遞網路(back propagation network, BPN)或學習向量量化(learning vector quantization, LVQ)網路，但在即時辨識的環境中，管制圖的形狀事實上是一種動態(dynamic)的時間序列(time series)，因此，問題的根本原因在於靜態類神經網路無法有效地辨識動態的管制圖形狀，然而訓練動態類神經網路（如Niocognitron）執行辨識任務是相當複雜且困難的。有鑑於此，本研究提出一個動態訓練法則，以提昇LVQ網路即時辨識管制圖異常形狀之績效。模擬結果顯示經過動態訓練之LVQ網路，在辨識管制圖異常形狀的精度與速度上，均優於此領域文獻中所提出的類神經網路辨識模式。雖然本研究是以提昇LVQ網路之辨識績效為主要的研究目的，本文所提出的動態訓練法則也可以應用在其他的類神經網路架構上，如自適應共振理論(adaptive resonance theory, ART)網路。

Unnatural control chart patterns (CCPs) are associated with a particular set of assignable causes for process variation. Hence, effectively recognizing CCPs can substantially narrow down the set of possible causes to be examined, and accelerate the diagnostic search. Recently, machine-learning techniques, especially the artificial neural network (ANN), have been widely used as an effective tool for CCP recognition (CCPR) tasks. Most ANN applications in CCPR have been using static supervised ANNs, such as back propagation networks (BPNs) and learning vector quantization (LVQ) networks. The false recognition problem (i.e. the patterns are misclassified) commonly encountered for these ANN-based CCPR models is mainly due to the fact that the static ANNs cannot appropriately deal with dynamic patterns, such as CCPs. In this research, a dynamic training algorithm is designed to provide an LVQ network-based CCPR model the capability to on-line recognize the dynamic CCPs that vary over time. The numerical results using simulation show that the dynamically trained LVQ network-based model proposed in this research performs much better than other ANN-based models reported in literature with respective to recognition accuracy and speed. Although this research considers the specific application of a real-time CCPR model based on an LVQ network, the proposed dynamic training algorithm could be applied to CCPR systems based on other ANN architectures in general.

1. Cheng, C. S.(1997).A neural network approach for the analysis of control chart patterns.International Journal of Production Research,35,667-697.
2. Cheng, C. S.,C. A. Tzeng(1994).A neural network approach for detecting shifts in the process mean and variability.Journal of the Chinese Institute of Industrial Engineers,11,67-75.
3. Chiu, C. C.,M. H. Yang(2001).Shifts recognition in correlated manufacturing process parameters using neural network.Journal of Chinese Institute of Industrial Engineers,18,86-94.
4. Davis, R. B.,W. H. Woodall(1988).Performance of the control chart trend rule under linear shift.Journal of Quality Technology,20,260-262.
5. Desieno, D.(1988).International Joint Conference on Neural Networks.San Diego, CA:IEEE.
6. Dieterle, F.,S. Muller-Hagedorn,H. M. Liebich,G. Gauglitz(2003).Urinary nucleosides as potential tumor markers evaluated by learning vector quantization.Artificial Intelligence in Medicine,28,265-279.
7. Duncan, A. J.(1986).Quality Control and Industrial Statistics.Illinois:Irwin Book Company.
8. Fukushima, K.,S. Miyake(1983).Neocognitron: a neural network model for a mechanism of visual pattern recognition.IEEE Transaction on Systems, Man, and Cybernetics,13,826-834.
9. Gan, F. F.(1992).CUSUM control charts under linear drift.The Statistician,41,71-84.
10. Gardner, J. W.,H. W. Shin,E. L. Hines,C. S. Dow(2000).An electronic nose system for monitoring the quality of potable water.Sensors and Actuators B,69,336-341.
11. Grant, L. E.,R. S. Leavenworth(1996).Statistical Quality Control.New York:McGraw-Hill Book Company.
12. Guh, R. S.(2005).A hybrid learning-based model for on-line detection and analysis of control chart patterns.Computers & Industrial Engineering,49,35-62.
13. Guh, R. S.(2002).Effects of non-normality on neural network based control chart pattern recognition.Journal of the Chinese Institute of Industrial Engineers,19,13-22.
14. Guh, R. S.,F. Zorriassatine,J. D. T. Tannock,C. O`Brien(1999).On-line control chart pattern detection and discrimination-A neural network approach.Artificial Intelligence in Engineering,13,413-425.
15. Guh, R. S.,J. D. T. Tannock(1999).Recognition of control chart concurrent patterns using a neural network approach.International Journal of Production Research,37,1743-1765.
16. Hassan, A.,M. Shariff Nabi Baksh,A. M. Shaharoun,H. Jamaluddin(2003).Improved SPC chart pattern recognition using statistical features.International Journal of Production Research,41,1587-1603.
17. Hwarng, H. B.(2005).Simultaneous identification of mean shift and correlation change in AR(1) processes.International Journal of Production Research,43,1761-11783.
18. Kohonen, T.(1995).Self Organization Maps.Heidelberg:Springer-Verlag.
19. Kohonen, T.(1988).An introduction to neural computing.Neural Networks,1,3-16.
20. Kohonen, T.,G. Barna,R. Chrisley(1988).International Joint Conference on Neural Networks.San Diego, CA:IEEE.
21. Law, A. M.,W. D. Kelton(1982).Simulation Modelling and Analysis.New York, NY:McGraw-Hill.
22. Lucas, J. M.(1982).Combined Shewhart-CUSUM Quality Control Schemes.Journal of Quality Technology,14,51-59.
23. Lucy-Bouler, T. L.(1993).Application to forecasting of neural network recognition of shifts and trends in quality control data.WCNN`93-World Congress on Neural Networks,1,631-633.
24. Lucy-Bouler, T. L.(1991).University of Alabama.
25. Medsker, L. R.(1995).Hybrid Intelligent Systems.Boston, MA, USA:Kluwer Academic.
26. Nelson, L. S.(1985).Interpreting Shewhart x-bar control charts.Journal of Quality Technology,17,114-116.
27. Nelson, L. S.(1984).The Shewhart control chart-tests for special causes.Journal of Quality Technology,16,237-239.
28. Noorossana, R.,M. Farrokhi,A. Saghaei(2003).Using neural networks to detect and classify out-of-control signals in autocorrelated processes.Quality and Reliability Engineering International,19,493-504.
29. Pearlmutter, B. A.(1989).Learning state space trajectories in recurrent neural networks.Neural Computation,1,263-269.
30. Pham, D. T.,E. Oztemel(1994).Control chart pattern recognition using learning vector quantization networks.International Journal of Production Research,32,721-729.
31. Rumelhart, D. E.,G. E. Hinton,R. J. Williams,D. E. Rumelhart,J. L. McClelland (eds)(1986).Parallel Distributed Processing: Explorations in the Microstructure of Cognition.Cambridge, MA:MIT press.
32. Shaffer, R. E.,S. L. Rose-Pehrsson,R. A. McGill(1999).A comparison study of chemical sensor array pattern recognition algorithms.Analytica Chimica Acta,384,305-317.
33. Specht, D. F.(1990).Probabilistic neural networks.Neural Networks,3,109-118.
34. Swift, J. A.,J. H. Mize(1995).Out-of-control pattern recognition and analysis for quality control charts using Lisp-based systems.Computers and Industrial Engineering,28,81-91.
35. Vuckovic, A.,V. Radivojevic,A. C. N. Chen,D. Popovic(2002).Automatic recognition of alertness and drowsiness from EEG by an artificial neural network.Medical Engineering and Physics,24,349-360.
36. Waibel, A.,T. Hanazawa,G. Hinton,K. Shikano,K. J. Lang(1989).Phoneme recognition using time-delay neural networks.IEEE Transactions on Acoustic, Speech and Signal Processing,37,328-339.
37. Wang, T. Y.,L. H. Chen(2002).Mean shifts detection and classification in multivariate process: a neural-fuzzy approach.Journal of Intelligent Manufacturing,13,211-221.
38. Western Electric.(1958).Statistical Quality Control Handbook.New York, NY:Western Electric Company.
39. Yang, M. S.,J. H. Yang(2002).A fuzzy-soft learning vector quantization for control chart pattern recognition.International Journal of Production Research,40,2721-2731.
40. Zhang, Z. P.,H. G. Chen,S. W. Ye,J. W. Zhao(1996).Comparison of the BP training algorithm and LVQ neural networks for e, μ, π identification.Nuclear Instruments and Methods in Physics Research A,379,271-275.
41. Zorriassatine, F.,R. S. Guh,R. M. Parkin,J. Coy(2004).Integrating novelty detection, neural networks and conventional tools for pattern recognition in multivariate processes.Journal of Engineering Manufacture, Proceedings of the Institute of Mechanical Engineers, Part B,218,779-793.