 |
|
|
|
以作者查詢圖書館館藏 、以作者查詢臺灣博碩士 、以作者查詢全國書目 、勘誤回報 、線上人數：32 、訪客IP：18.117.7.48
姓名 林湧清(Yung-Cing Lin) 查詢紙本館藏 畢業系所 機械工程學系 論文名稱 基於人工智慧之晶圓瑕疵檢測採樣策略研究
(Artificial Intelligent Based Sampling Strategy for Wafer Defect Inspection)相關論文 檔案 [Endnote RIS 格式]
[相關文章]
[文章引用]
[完整記錄]
[館藏目錄]
至系統瀏覽論文 (2029-8-1以後開放)
摘要(中) 在半導體製造過程中，確保晶圓的品質對於電子設備的性能和可靠性至
關重要。傳統的瑕疵採樣策略，如靜態和動態採樣，通常無法應對瑕疵檢測
中不斷變化的複雜性，導致檢測效率低下且人力成本高昂。為了解決這一問
題，本研究提出了一種利用機器學習技術的先進晶圓瑕疵檢測策略。該策略
採用了隨機森林(RF)和極限梯度提升(XGBoost)模型，結合合成少數過採樣
技術(SMOTE)，並使用基於密度的空間聚類（DBSCAN）方法，將取樣對象
分為非群聚瑕疵和群聚瑕疵，從而提高瑕疵辨識的精確度和效率。對於非群
聚性瑕疵，使用改良式取樣增益圖來探討所需辨識瑕疵的機率閥值，對於群
聚性瑕疵，根據群聚內晶粒為瑕疵的機率做排序，取樣瑕疵機率最高的前五
點做人工複檢。
實驗證明，本研究所提出的採樣策略顯著降低了缺陷檢測點數，同時保
持了高水準的瑕疵召回率。對於非群聚瑕疵，當對瑕疵機率超過 20%的晶
粒進行取樣時，非群聚瑕疵的召回率達到 90%，顯示所提出的策略能有效
識別散佈於晶圓上的個別瑕疵。此外，對於群聚瑕疵，所提出的策略在幾乎
沒有瑕疵誤報率的情況下，以極少的取樣點數成功辨識了群聚內所有瑕疵。
對比於合作半導體廠商所開發的採樣演算法，本研究提出基於機器學習的
採樣策略在達到 90%瑕疵召回率的前提下，可以降低近 80%的總採樣點數。
這不僅顯著提升了檢測效率，還大幅減少了生產成本。因此，該採樣策略比
傳統方法更具效率，並且已具備在產線上實際應用的價值。摘要(英) Ensuring the quality of wafers in semiconductor manufacturing is crucial for the performance and reliability of electronic devices. Traditional defect sampling strategies, such as static and dynamic sampling strategy, often fail to cope with the increasing complexity of defect detection, leading to inefficiencies and high labor costs. To address this issue, this study proposes an advanced wafer defect sampling strategy leveraging machine learning techniques. The proposed strategy employs Random Forest (RF) and Extreme Gradient Boosting (XGBoost) models, combined with the Synthetic Minority Over-sampling Technique (SMOTE), and utilizes the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) method to classify defects into non-clustered and clustered categories, thereby enhancing the accuracy and efficiency of defect identification. For non-clustered defects, the Modified Sampling Gains Chart is used to determine the probability threshold required for defect identification. For clustered defects, the defects are ranked based on the probability of the dies within the cluster being defective, and the top five dies with the highest defect probability are sampled.
Validation experiments demonstrate that the proposed strategy significantly reduces the number of sampling dies while maintaining a high recall rate of defect. For non-clustered defects, when sampling dies with a defect probability exceeding 20%, the recall rate for nonclustered defects reaches 90%. This indicates that the proposed strategy effectively identifies individual defects across the wafer. Additionally, for clustered defects, the proposed strategy successfully identifies all defects within the clusters with minimal false positives, using a low number of sampling points. Compared to the sampling algorithms developed by the collaborative semiconductor manufacturer, the machine learning-based sampling strategy proposed in this study achieves a 90% defect recall rate while reducing the total number of sampling points by nearly 80%. This not only significantly enhances detection efficiency but also substantially reduces production costs. Consequently, this sampling strategy is more efficient than traditional methods and demonstrates practical applicability in production lines.關鍵字(中) ★ 人工智慧
★ 聚類演算法
★ 晶圓瑕疵採樣策略關鍵字(英) ★ Artificial Intelligent
★ Clustering Algorithms
★ Wafer Defect Sampling Strategy論文目次 ABSTRACT .............................................................................................................................Ⅰ
ACKNOWLEDGEMENTS .....................................................................................................IV
TABLE OF CONTENTS...........................................................................................................V
LIST OF TABLES.................................................................................................................. VII
LIST OF FIGURES..................................................................................................................IX
1. INTRODUCTION...............................................................................................................1
1.1 Semiconductor Wafer Defect Detection.....................................................................1
1.2 Traditional Defect Sampling Strategies.....................................................................2
1.3 Application of Machine Learning in Wafer Defect Detection ...................................3
1.4 Purpose.......................................................................................................................8
2. METHOD AND THEORY.................................................................................................9
2.1 Overview of Machine Learning Models....................................................................9
2.1.1 Random Forest (RF).......................................................................................9
2.1.2 Logistic Regression (LR).............................................................................10
2.1.3 XGBoost.......................................................................................................12
2.2 Density-Based Spatial Clustering of Applications with Noise (DBSCAN) ............15
2.3 Over-Sampling and Under-Sampling Methods........................................................17
2.4 Model Hyperparameter Optimization Method - Particle Swarm Optimization
Algorithm ................................................................................................................20
2.5 Dataset from a Semiconductor Manufacturer and It’s Current Defect Sampling
Strategy....................................................................................................................22
2.6 Proposed Wafer Defect Sampling Strategy..............................................................25
2.7 Cross-Validation Method .........................................................................................26
2.8 Modeling Procedure.................................................................................................27
VI
3. RESULTS AND DISCUSSION .......................................................................................30
3.1 Comparison of Performance of Different Machine Learning Models.....................30
3.2 Effect of DBSCAN on Defect Sampling Strategy ...................................................37
3.3 Proposed Sampling Strategy Applied to a Real-world Scenario .............................40
4. CONCLUSIONS...............................................................................................................51
REFERENCES.........................................................................................................................52
APPENDIX. SUPPLEMENTARY MATERIALS ..................................................................55參考文獻 1. S. K. Fan, C. Y. Lin, C. H. Jen, K. L. Chen, and L. T. Jung, “Defective Wafer Detection Using a Denoising Autoencoder for Semiconductor Manufacturing Processes,” Advanced Engineering Informatics, Vol. 46, 101178, 2020.
2. L. Milor, “A Survey of Yield Modeling and Yield Enhancement Methods,” IEEE Transactions on Semiconductor Manufacturing, Vol. 26, No. 2, pp. 196-213, 2013.
3. L. Li, H. L. Yan, J. L. Li, Q. L. Li, T. Zhu, H. Wu, R. Liu, R. Jin, and J.M. Wu, “Effect of Wafer Defects on Electrical Properties and Yields of SiC Devices,” Journal of Physics: Conference Series, Vol. 2033, 012095, 2021.
4. R. K. Nurani, R. Akella, A. J. Strojwas, R. Wallace, M. G. Mclntyre, J. Shields, and I. Emami, “Development of an Optimal Sampling Strategy for Wafer Inspection,” Proceedings of International Symposium on Semiconductor Manufacturing, Tokyo, Japan,pp. 143-146, 1994.
5. S. Dauzère-Pérès, J. L. Rouveyrol, C. Yugma, and P. Vialletelle, "A Smart Sampling Algorithm to Minimize Risk Dynamically," 2010 IEEE/SEMI Advanced Semiconductor Manufacturing Conference (ASMC), San Francisco, CA, USA, pp. 307-310, 2010.
6. C. W. Liu and J. Ou, "Use Machine Learning Based Smart Sampling to Improve System Level Testing Efficiency," 2021 IEEE International Test Conference in Asia (ITC-Asia), Shanghai, China, pp. 1-6, 2021.
7. T. H. Lee, K. B. Lee, and C. O. Kim, "Performance of Machine Learning Algorithms for Class-Imbalanced Process Fault Detection Problems," IEEE Transactions on
Semiconductor Manufacturing, Vol. 29, No. 4, pp. 436-445, 2016.
8. S. K. Fan, C. W. Cheng, and D. M. Tsai, "Fault Diagnosis of Wafer Acceptance Test and Chip Probing Between Front-End-of-Line and Back-End-of-Line Processes," IEEE Transactions on Automation Science and Engineering, Vol. 19, No. 4, pp. 3068-3082, 2022.
9. T. K. Ho, "Random Decision Forests," Proceedings of 3rd International Conference on Document Analysis and Recognition, Montreal, QC, Canada, Vol. 1, pp. 278-282, 1995.
10. T. K. Ho, "The Random Subspace Method for Constructing Decision Forests," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 20, No. 8, pp. 832-844, 1998.
11. L. Breiman, "Bagging Predictors," Machine Learning, Vol. 24, pp. 123–140, 1996.
12. L. Breiman, "Random Forests," Machine Learning, Vol. 45, pp. 5–32, 2001.
13. C. Y. Peng, K. L. Lee, and G. M. Ingersoll, "An Introduction to Logistic Regression Analysis and Reporting," The Journal of Educational Research, Vol. 96, No.1, pp. 3-14,
2002.
14. C. G. Fan, Z. W. Xie, Y. W. Liu, C. Y. Li, and H. Liu, " Adaptive Controller Based on Spatial Disturbance Observer in a Microgravity Environment," Sensors, Vol. 19, No. 21,4759, 2019.
15. Medium, #11 What is a Linear Classifier (Logistic Regression),https://medium.com/@musicaround/11-what-is-a-linear-classifier-logistic-regression4eb44e2544b4, accessed on March 8, 2024.
16. T. Q. Chen and C. Guestrin, "XGBoost: A Scalable Tree Boosting System," Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, pp. 785-794, 2016.
17. R. Guo, Y. T. Liu, Z. Q. Zhao, J. Y. Zhao, J. W. Wang, and W. Cai, "Research on Degradation State Recognition of Axial Piston Pump under Variable Rotating Speed," Processes, Vol. 10, No. 6, 1078, 2022.
18. J. R. Quinlan, "Induction of Decision Trees," Machine Learning, Vol. 1, pp. 81-106, 1986.
19. M. Ester, H. P. Kriegel, J. Sander, and X. W. Xu, "A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise," KDD′96: Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, Portland, Oregon, pp. 226-231, 1996.
20. T. Falkowski, Community Analysis in Dynamic Social Networks, Ph.D. Thesis, Otto von Guericke University, Magdeburg, Germany, 2009.
21. N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, "SMOTE: Synthetic Minority Over-sampling Technique," Journal of Artificial Intelligence Research, Vol. 16, pp. 321-357, 2002.
22. Rich Data, SMOTE Explained for Noobs – Synthetic Minority Over-sampling Technique Line by Line, https://rikunert.com/smote_explained, accessed on March 8, 2024.
23. I. Tomek, "Two Modifications of CNN," IEEE Transactions on Systems, Man, and Cybernetics, Vol. 6, pp. 769–772, 1976.
24. Imbalanced Learn, User Guide, https://imbalancedlearn.org/stable/under_sampling.html#, accessed on March 8, 2024.
25. J. Kennedy and R. Eberhart, "Particle Swarm Optimization," Proceedings of ICNN′95 - International Conference on Neural Networks, Perth, WA, Australia, Vol. 4, pp. 1942-1948,1995.
26. Scikit-Learn, Cross-validation: Evaluating Estimator Performance, https://scikitlearn.org/stable/modules/cross_validation.html, accessed on March 28, 2024指導教授 林志光(Chih-Kuang Lin) 審核日期 2024-7-18 推文 plurk
funp
live
udn
HD
myshare
netvibes
friend
youpush
delicious
baidu
網路書籤 Google bookmarks
del.icio.us
hemidemi
facebook
twitter
google
reddit