跳到主要內容

臺灣博碩士論文加值系統

(35.172.136.29) 您好!臺灣時間:2021/07/25 01:53
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

我願授權國圖
: 
twitterline
研究生:黃識翰
研究生(外文):Shih-Han Huang
論文名稱:利用獨立成分分析進行半導體製程錯誤訊號偵測之研究
論文名稱(外文):Fault detection Using Independent Component Analysis in Semiconductor Processes
指導教授:范書愷范書愷引用關係
口試委員:黃乾怡邱垂昱蔡篤銘
口試日期:2012-06-12
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:工業工程與管理系碩士班
學門:工程學門
學類:工業工程學類
論文種類:學術論文
論文出版年:2012
畢業學年度:100
語文別:英文
論文頁數:82
中文關鍵詞:製程監控獨立成分分析半導體製造錯誤偵測
外文關鍵詞:Process monitoringIndependent Component Analysis (ICA)Semiconductor manufacturingFault detection
相關次數:
  • 被引用被引用:0
  • 點閱點閱:150
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
在半導體產業邁向高度自動化的今日,製程操作大都擁有數量龐大及種類繁多的操作數據資料,因此發展一套製程錯誤偵測方法,察覺製程狀態是否發生異常是極為重要的任務。而半導體製程的偵測,必須透過量測器(sensor)取得製程變數及製程輸出的正確訊息,才能了解製程異常及錯誤(fault)發生的精確時間及其型態。然而,量測器會發生時間延遲 (time delay) 與資料佚失 (missing data)等問題,因此,上述問題對錯誤偵測將產生巨大的影響。本研究主要以濺鍍製程為例,我們採用四種製程參數分別為Gas2、Gas3、PRESSURE以及PWR,在此我們以獨立成分分析來解析製程資料,再將由獨立成分分析所得到的基底資料建立為製程錯誤偵測的基礎,並依據過去操作資訊建立製程監控的模型。

Nowadays, semiconductor industry has been marching toward an increasingly automated, ubiquitous data gathering production system that is full of manufacturing complexity and environment uncertainty. Therefore developing an effective fault detection system to reflect the current operation status is essential. In order to facilitate the fault diagnosis and tally the time of fault really occurred, the process engineer must collect the correct information of process variables and process outputs from the sensor. However, times delay and missing values in the sensor occur quite often in practice. In this research we focus on the physical vapor deposition (PVD) and we have four process parameter that include Gas2, Gas3, PRESSURE and PWR. Thus, in order to rectify the aforementioned difficulties that could realistically take place in practice, we use the Independent Component Analysis to decompose each process parameter into the basis data. An effective method is presented to identify the faults of the process and construct a process monitoring model by means of these basis signals.

摘 要 i
Abstract ii
List of Tables vi
List of Figures vii
Chapter1 INTRODUCTION 1
1.1 Background 1
1.2 Motivation 2
1.3 The Research Objectives 3
1.4 Organization of The Thesis 4
Chapter2 LITERATURE REVIEW 5
2.1 The Flexible Golden Pattern 5
2.1.1 Definition of Input Data 6
2.1.2 Normalization 6
2.1.3 Score Algorithm 7
2.2 Independent Component Analysis (ICA) 9
2.2.1 Introduction of ICA 9
2.2.2 Mathematical Definitions of ICA 10
2.2.3 Definition of Statistical Independence 11
2.2.4 Maximum Non-Gaussianity Versus Statistical Independence 12
2.3 Principal Component Analysis 15
2.4 Statistical Process Control 17
2.4.1 Shewhart Control Chart 17
2.4.2 CUSUM Control Chart 18
2.4.3 EWMA Control Chart 19
2.5 The Hotelling T Control Chart 20
Chapter 3 SIGNAL DETECTION USING INDEPENDENT COMPONENT ANALYSIS 23
3.1 Time Delay and Missing Values 23
3.1.1 Time Delay 24
3.1.2 Missing Values 25
3.2 ICA Method for Fault Detection 28
3.3 Independent Component Analysis 31
3.3.1 ICA Algorithm 32
3.3.2 Preprocessing ICA 33
3.3.3 FastICA Algorithm 35
3.4 Training The Basis Data and Reconstructing The Testing Data 39
3.4.1 Training The Basis Data 39
3.4.2 Reconstructing The Testing Data 40
Chapter 4 EXPERIMENTAL RESULTS 42
4.1 Introduction of Sputter Deposition 42
4.2 Experimental Results of ICA 43
4.3 Performance Comparison with Hotelling 54
4.4 Experimental Results of Dealing with Time Delay and Missing Values 57
4.4.1 Experimental Results of Dealing with Time Delay 58
4.4.2 Experimental Results of Dealing with Missing Values 60
Chapter 5 CONCLUSIONS 68
REFERENCES 71
Appendix A 75
Appendix B 81



[1] W. Bernhard, “Handbook of ion sources,” CRC Press, vol. 14, no.6, 1995, p. 222.
[2] B. A. Movchan and A. V. Demchishin, “Study of the structure and properties of thick vacuum condensates of nickel, titanium, tungsten, aluminium oxide and zirconium dioxide,” Phys. Met. Metallogr, vol. 28, 1969, pp. 83-90.
[3] Y. Ding, L. Zeng and S. Zhou, “Phase I Analysis for Monitoring Nonlinear Profiles in Manufacturing Processes,” Journal of Quality Technology, vol. 38, no. 3, 2006, pp. 199-216.
[4] H. Hotelling, “Multivariate Quality Control,” Techniques of Statistical Analysis, vol 25, no. 6, 1947, pp. 111-184.
[5] D. S. Holmes and A. E. Mergen, “Improving the performance of the Control Chart,” Quality Engineering, vol. 5, no. 10,1993, pp. 619-625.
[6] H. Windischman, “Itrinsic stress in sputter-deposited thin film,” Crit. Rev. Sol. St. Mat. Sci, vol. 17, no. 6, 1992, pp. 547.
[7] J. Hurri, H. Gavert, J. Sarela and A. Hyvarinen, FastICA Package, 2004, Online Available: http://www.cis.hut.fi/projects/ica/fastica/.
[8] A. Hyvarinen, J. Karhunen and E. Oja,, Independent Component Analysis, New York: John Wiley & Sons, 2001.
[9] A. Hyvarinen, and E. Oja., “A fast fixed-point algorithm for independent component analysis,” Neural Computation, vol. 9, no. 8, 1991, pp. 1483-1492.
[10] J. Jin and J. Shi, “Feature-Preserving Data Compression of Stamping Tonnage Information Using Wavelets,” Technometrics, vol. 41, no. 4, 1999, pp. 327-339.
[11] J. L. Smith, G. M. Russel and S.C. Bhatia, “Heat Treatment of Metal Vol.2,” Alkem Oxford, vol. 2, no. 25, 2009, pp. 95-97.
[12] J.A. Thornton, “Influence of apparatus geometry and deposition conditions on the structure and topography of thick sputtered coatings,” J. Vac. Sci. Tech, vol. 11, no. 23, 1974, pp. 666-670.
[13] L. Kang and S. L. Albin, “On-Line Monitoring When the Process Yields a Linear Profile,” Journal of Quality Technology, vol. 32, no. 4, 2000, pp. 418-426.
[14] K. Kim, M. A. Mahmoud and W. H. Woodall, “On the Monitoring of Linear Profile,” Journal of Quality Technology, vol. 35, no. 3, 2003, pp. 317-328.
[15] K. Lshii, “High-rate low kinetic energy gas-flow-sputtering system,” J. Vac. Sci. Technol. A, vol. 7, no. 15, 1989, pp. 256-258.
[16] K. Ortner, M. Birkholz and T. Jung, “Neue Entwicklungen beim Hohlkatoden-Gasflusssputtern,” Vac. Praxis, vol. 15, no. 21, 2003, pp. 236-239.
[17] M. A. Mahmoud and W. H. Woodall, “Phase I analysis of Linear Profiles with Calibration Applications,” Technometrics, vol. 46, no. 15, 2004, pp. 377-391.
[18] J. M. Moguerza, A. Munoz and S. Psarakis, “Monitoring Nonlinear Profiles Using Support vector Machines,” Lecture Notes in Computer Science , vol. 4789, no. 31, 2007 , pp. 574-583.
[19] D. C. Montgommery, Introduction to Statistical Quality Control, 5th Edition, New York: Wiley, 2005.
[20] O. Mestek, J. Pavlik and M. Suchanek, “Multivariate Control Charts: Control Charts for Calibration Curves,” Fresenius Journal of Analytical Chemistry, vol. 350, no. 10, 1994, pp. 344-351.
[21] D. C. Montgomery, Introduction to Statistical Quality Control, 5th Edition, New York: Wiley, 2005, pp. 11-36.
[22] M. Birkholz, C. Genzel and T. Jung, “X-ray diffraction study of residual stress and preferred orientation in thin titanium films subjected to a high ion flux during deposition,” J. Appl. Phys, vol. 96, no. 24, 2004, pp. 7202-7211.
[23] J. S. Oakland, Statistical Process Control, Fourth Edition, Butterworth-Heinemann: Oxford, 1999, pp. 14-18.
[24] R. E. Quandt, “The Estimation of the Parameters of a Linear Regression System Obeying Two Separate Regimes,” Journal of the American Statistical Association, vol. 53, no. 35, 1958, pp. 873-880.
[25] T. Jung and A. Westphal, “Zirconia thin film deposition on silicon by reactive gas flow sputtering: the influence of low energy particle bombardment,” Mat. Sc. Eng. A, vol. 140, no. 8, 1991, pp. 528-533.
[26] H. W. WILLIAM, “Controversies and Contradictions in Statistical Process Control,” Journal of Quality Technology, vol. 32, no. 4, 2000.
[27] K. Wang and F. Tsung, “Using Profile Monitoring Techniques for a Data-Rich Environment with Huge Sample Size,” Quality and Reliability Engineering International, vol. 21, no. 16, 2005, pp. 677-688.
[28] W. H. Woodall, D. J. Spitzner, D. C. Montgomery and S. Gupta, “Using Control Charts to Monitor Process and Product Quality Profiles,” Journal of Quality Technology, vol. 36, no. 3, 2004, pp. 309-320.
[29] S. H. Wu, Machine vision-based defect detection of solar cells/modules in electroluminescence images, Master thesis , Yuan-Ze university, department of industrial engineering, 2011.
[30] C. P. Quesenberry, “SPC Methods for Quality Improvement,” John Wiley & Sons, INC, vol. 35, no. 27, 1997, pp. 30-31
[31] W. H. Woodall and B. B. Jeffrey, “Destribution of Hotelling Statistic Based on the Sucessive Differences Estimator,” Joe H Sullivan Journal of Quality Technology, vol. 11, no. 11, 2006, pp. 217.
[32] J. D. Williams, W. H. Woodall, J. B. Birch and J. H. Sullivan,“Distribution of Hotelling’s based on successive differences Covariance Matrix Estimator,” Journal of Quality Technology, vol. 38, no. 3, 2006, pp. 217-229.
[33] J. D. Williams, W. H. Woodall and J. B. Birch, “Statistical Monitoring of Nonlinear Product and Process Quality Profiles,” Quality and Reliability Engineering International, vol. 23, no. 8, 2007, pp. 925-941.
[34] W. H. Woodall, D. J. Spitzner, D. C. Montgomery and S. Gutpa, “Using Control Charts to Monitor Process and Product Quality profiles,” Journal of Quality Technology, vol. 36, no. 6, 2004, pp. 309-320.
[35] K. Wang and F. Tsung, “Using Profile Monitoring Techniques for a Data-Rich Environment with Huge Sample Size,” Quality and Reliability Engineering International, vol. 21, no. 4, 2005, pp. 677-688.


QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top