博碩士論文 954206001 詳細資訊


姓名 楊政霖(Cheng-Lin Yang)  查詢紙本館藏   畢業系所 工業管理研究所
論文名稱 應用以田口方法最適化之倒傳遞類神經網路於TFT-LCD Cell製程缺陷分類之研究
(Optimum Design for BPNN with Taguchi Method to Study on Defects Classification in TFT-LCD Cell Process)
檔案 [Endnote RIS 格式]    [Bibtex 格式]    至系統瀏覽論文 ( 永不開放)
摘要(中) TFT-LCD是目前應用最為廣泛之平面顯示器,然而其生產製造過程卻不如銷售般那樣順遂。通常TFT-LCD的誕生是需要經過多道繁瑣的製造程序,因此每當產品有不良缺陷產生時,便很難在短時間內斷定其不良缺陷的成因。此外,這些不良缺陷的分析常仰賴有經驗的工程師,對於新進的職員而言,短時間學會如何分辨不良缺陷著實是一項艱難的任務。從管理層面的觀點來看,訓練及解析問題所耗之時間與成本顯然不符合成本經濟效益,因此,發展或建立一套缺陷分類系統或模型是勢在必行的。
本研究經由文獻之探討,發現近年來倒傳遞類神經網路已成功地應用在分類及預測領域上,因此本研究經仔細評估與判斷後,決定採用倒傳遞類神經網路來研究在TFT-LCD Cell製程缺陷分類之問題。為了使網路的表現最適化,本研究採用田口實驗來決定關鍵性之參數,進而選出最適之參數組合。
根據數值分析之結果顯示網路參數-轉換函數、學習速率及世代大小之主效應及彼此之交互作用對於網路的表現皆有顯著性之影響,因此在選擇最適參數組合時,需透過二維或一維之參數平均表現反應表或Duncan’s Multiple Range Test來做決定。最後,將最適參數組合應用在倒傳遞類神經網路中,提供給廠商做為未來缺陷分類之依據。
摘要(英) TFT-LCDs have become one of the most popular flat panel display devices in these days and applied to various fields in the world. However, manufacturing the TFT-LCD panel indeed requires passing through many complex processes. It would be a tough task to analyze the causes of defects when the defects occur. As a result, manufactures need an efficient method to help employees quickly clear the causes of defects and then do the appropriate treatments to avoid resulting in a mass profit loss.
In this study, we use BPNN to approach the relationship between defective characteristics and causes of defects so as to solve the defects classification problem in the TFT-LCD cell process. In order to make BPNN perform well, we adopt Taguchi method to study not only the significant parameters but the optimum parameter settings in BPNN.
Numerical analysis results show the main and interaction effects of parameters- transfer function, learning rate, and epoch size have significant effects on the performance of BPNN in our case. Besides, the corresponding optimal levels of each significant parameter can be determined by two-way, one-way table, or Duncan’s Multiple Range Test. At last, we extract the final weights and biases from trained BPNN at the optimal condition and then provide this network model with the optimum designs for manufacturers to apply to defects classification problem in TFT-LCD cell process.
關鍵字(中) ★ 直交表
★ 田口方法
★ 倒傳遞類神經網路
★ 缺陷分類
關鍵字(英) ★ Taguchi method
★ BPNN
★ Orthogonal array
★ Defect classification
論文目次 摘要 i
Abstract ii
誌謝 iii
Table of Content iv
List of Figures vi
List of Tables vii
Chapter 1 Introduction 1
1.1 Research motivation and background 1
1.2 Problem description 4
1.3 Research objectives 5
1.4 Research methodology and framework 6
1.4.1 Research methodology 6
1.4.2 Research framework 7
Chapter 2 Literature review 9
2.1 Defect classification 9
2.2 Back-propagation neural network 11
2.3 Taguchi Method 15
Chapter 3 Research methodology 19
3.1 Data processing 19
3.2 Parameters selection 20
3.3 Taguchi experiment 23
3.3.1 Identification of performance characteristic and design variables 24
3.3.2 Determination of the levels and an orthogonal array 25
3.3.3 Experimentation 26
3.3.4 Experiment analysis 31
3.3.5 Confirmation 34
Chapter 4 Numerical analysis 36
4.1 Data setting 36
4.2 Taguchi experiment 42
4.2.1 Experiment setting 43
4.2.2 Experimentation 47
4.2.3 Results and analysis 48
4.3 Results discussion 55
Chapter 5 Conclusion and future work 57
5.1 Conclusion 57
5.2 Future work 58
Reference 59
參考文獻 [1] Bishop, C.M. (1995), Neural networks for pattern recognition. Clarendon Press, Oxford, UK.
[2] Bounds, D.G., Lloyd, P,J. (1988), A multilayer perceptron network for the diagnosis of low back pain. Proc. Second IEEE Int'l. Conf. Neural Networks, San Diego, July 24-27, II-481-II-489.
[3] Castillo, P.A., Merelo, J.J., Prieto, A., Rivas, V., Romero, G. (2000), Evolving multilayer perceptrons. Neural Processing Letters 12, 115-127.
[4] Castillo, P.A., Merelo, J.J., Prieto, A., Rivas, V., Romero, G. (2000), G-Prop: Global optimization of multilayer perceptrons using GAs. Neurocomputing 35, 149-163.
[5] Chu, C.Y., The research of defect solutions for TFT-LCD G4.5 cell process in BPN application, National Central University, Executive Master of Industrial Management, 2007.
[6] Cybenko,G. (1989), Approximation by superpositions of a sigmoidal function. Mathematics of Control, Signals and Systems 2(4), 303-314.
[7] Dayhoff, Judith E. (1990) Neural network architectures: an introduction. Van Nostrand Reinhold, New York.
[8] Feng, C.X.J., Gowrisankar, A.C., Smith, A.E., Yu, Z.G.S. (2006), Practical guidelines for developing BP neural network models of measurement uncertainty data. Journal of Manufacturing System 25(4), 239-250.
[9] Ham, J., Kamber, M. (2003) Data mining: concepts and techniques. Morgan Kaufmann, San Francisco, California.
[10] Hart, A. (1992), Using neural networks for classification tasks-some experiments on datasets and practical advice. Journal of the Operational Research Society 43(3), 215-226.
[11] Haykin, S. (1994), Neural networks: A Comprehensive foundation. Prentice-Hall International, Englewood Cliffs, NJ.
[12] Hsieh, K.L., Lu, Y.S. (2008), Model construction and parameter effect for TFT-LCD process based on yield by using ANNs and stepwise regression. Expert Systems with Application 34, 717-724.
[13] Khaw, J.F.C, Lim, B.S., Lim, L.E.N. (1995), Optimal design of neural network using the Taguchi method. Neurocomputing 7, 225-245.
[14] Kim, Y.S., Yum, B.J. (2004), Robust design of multilayer feedforward neural networks: an experimental approach. Engineering Applications of Artificial Intelligence 17, 249-263.
[15] Lee, K.H, Yi, J.W., Park, J.S., Park, G.J. (2003), An optimization algorithm using orthogonal arrays in discrete design space for structures. Finite Elements in Analysis and Design 40, 121-135.
[16] Leonard, J.A., Kramer, M.A. (1991), Radial basis function networks for classifying process faults. Control Systems Magazine 11(3), 31-38.
[17] Lim, D.C., Seo, D.G., Jeong, D.H. (2005), Defect Classification for Inspection of TFT-LCD Glass. Proceedings of SPIE 6051, 60510F-1-60510F-6.
[18] Lin, S.W., Chou, S.Y., Chen, S.C. (2007), Irregular shapes classification by back-propagation neural networks. International Journal of Advance Manufacturing Technology 34, 1164-1172.
[19] Lin, S.W., Tseng, T.Y., Chou, S.Y., Chen, S.C. (2008), A simulated-annealing-based approach for simultaneous parameter optimization and feature selection of back-propagation networks. Expert Systems with Application 34, 1491-1499.
[20] Lin, T.Y., Tseng, C.H. (2000), Optimum design for artificial neural networks: an example in a bicycle derailleur system. Engineering Application of Artificial Intelligence 13, 3-14.
[21] Maier, H.R., Dandy, G.C. (1998), The effect of internal parameters and geometry on the performance of back-propagation neural networks: an empirical study. Environmental Modelling & Software 13, 193-209.
[22] Maren, A., Harston, C., Pap, R. (1990), Handbook of Neural Computing Applications, Academic Press, San Diego, CA.
[23] Montgomery, D.C. (1997), Design and Analysis of Experiments, 5th Edition, Wiley, New York.
[24] NeuralWare Inc. (1991). Neural Computing, NeuralWorks Professional II/Plus and NeuralWorks Explorer.
[25] Packianather, M.S., Drake, P.R., Rowlands, H. (2000), Optimizing the parameters of multilayered feedforward neural networks through Taguchi design of experiments. Quality and Reliability international 16, 461-473.
[26] Park, S.H. (1996), Robust Design and Analysis for Quality Engineering, Chapman & Hall, London.
[27] Peace, G.S. (1993), Taguchi Method: A Hands-on Approach, Addison-Wesley, Reading, MA.
[28] Pham, D.T., Sagiroglu, S. (2000). Neural network classification defects in veneer boards. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture 214(3), 255-258.
[29] Sexton, R. S., Alidance, B., Dorsey, R. E. (1998). Global optimization for artificial neural network: a Tabu search application. European Journal of Operational Research 106, 570-584.
[30] Sukthomya, W., Tannock, J. (2005), The optimization of neural network parameters using Taguchi's design of experiments approach: an application in manufacturing process modeling. Neural Computing and Application 14, 337-344.
[31] Taguchi, G. (1987), System of Experimental Design, Vol. 1& 2. UNIPUB/Kraus International Publications, New York.
[32] Tortum, A., Yayla, N., Celik, C., Gokdag, M. (2007), The investigation of model selection criteria in artificial neural networks by Taguchi method. Physica A 386, 446-468.
[33] Wang, T.Y., Huang, C.Y. (2008), Optimizing back-propagation networks via a calibrated heuristic algorithm with an orthogonal array. Expert Systems with Application 34, 1630-1641.
[34] Yang, T., Lin, H.C., Chen, M.L. (2006), Metamodeling approach in solving the machine parameters optimization problem using neural network and genetic algorithms: A case study. Robotics and Computer-Integrated Manufacturing 22, 322-331.
[35] Yang, T., Olmen, R.V. (2004), Robust design for a multilayer ceramic capacitor screen-printing process case study. Journal of Engineering Design 15(5), 447-457.
[36] Yuen, C.W.M., Wong, W.K., Qian, S.Q., Chan, L.K., Fung, E.H.K. (2008), A hybrid model using genetic algorithm and neural network for classifying garment defects. Expert Systems with Applications.
指導教授 沈國基(Gwo-Ji Sheen) 審核日期 2008-7-14
推文 facebook   plurk   twitter   funp   google   live   udn   HD   myshare   reddit   netvibes   friend   youpush   delicious   baidu   

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡