使用聚類過濾策略和 CNN 計算識別晶圓圖瑕疵樣態

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：33

、訪客IP：3.143.235.212

姓名

黃韋澄(Wei-Cheng Huang) 查詢紙本館藏

畢業系所

電機工程學系

論文名稱

使用聚類過濾策略和 CNN 計算識別晶圓圖瑕疵樣態
(Wafer Defect Pattern Identification Using Cluster Filtering Strategy and CNN Computation)

相關論文

★ 晶圓圖提取特徵參數錯誤樣態分析

★ 新建晶圓圖相似性門檻以強化相似程度辨別能力

檔案

[Endnote RIS 格式]

[Bibtex 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 ( 永不開放)

摘要(中)

隨著製程的日新月異，使得電晶體不斷的縮小，得到了運算速度提升、消耗功率降低等諸多好處，但同時產生了晶圓良率的下降，在探針測試(Probe testing)階段會對所有晶粒做電性測試，未能通過此檢測的晶粒會被標記為壞晶粒，最後形成整張晶圓圖，我們能透過晶圓圖上壞晶粒所形成的形狀，來找出特定的製程錯誤，過去以人工對晶圓圖瑕疵樣態做識別，雖然能直覺的進行分類，但也耗費了大量時間與人力，為了解決上述的問題，我們希望以一套自動化系統去對晶圓圖做瑕疵樣態的識別。本論文可以分為兩個階段，第一階段使用連通分量分析對壞晶粒進行分群處理，再將晶圓圖進行圓形切割，分為內、中、外三個區域計算其特徵參數，我們針對每一個晶圓圖的特徵參數選擇合適的群聚濾波器，濾除掉對顯著樣態沒有太大影響的離群晶粒，第二階段則利用深度學習中的卷積神經網路將晶圓圖分為九種瑕疵樣態，我們的方法運算時間為每張晶圓圖17.56毫秒，八種樣態精度可達到96.03%，並給予原先定義無樣態的None一個參考。

摘要(英)

Due to the progress of manufacturing process, transistor size has reduced but circuit size has increased nowadays. Defects on a wafer have increased, affecting the yield and increasing the cost. During probe test on all dies of a wafer, dies failing to pass test are marked as bad dies. Defect pattern wafer maps are constructed by these bad dies. We can find out specific process errors by analyzing defect patterns, especially efficient if using an automatic system to analyze wafer maps.
In the first stage of this thesis, we preprocess wafer maps by using connected component analysis to group bad dies and calculating the average clustering parameters. We carry out circular segmentation for each wafer map, i.e. dividing the map into inner, middle and outer regions, and analyze their characteristics. We obtain the characteristic parameters such as edge bad die proportion, density of bad dies and BD focus, among which the last one represents the key distribution area of bad dies. We classify wafer maps according to the key distribution of bad dies and formulate a filter selection strategy to filter out suitable outliers to obtain required clusters. In the second stage we use CNN for the remaining clusters to classify nine defect patterns on the processed wafer maps.
The accuracy of our experiment is 85.36% for nine types of defect patterns, and the accuracy of eight types other than None reaches 96.03%. Our method can classify those maps that may not easy to be recognized. The average speed of our method is 17.56 milliseconds per wafer map.

關鍵字(中)

★ 晶圓圖
★ 錯誤樣態辨識
★ 良率
★ 群聚

關鍵字(英)

★ Wafer map
★ Defect pattern recognition
★ Yield
★ Cluster

論文目次

中文摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
圖目錄 vi
表目錄 viii
第一章緒論 1
1-1 前言 1
1-2 研究動機 3
1-3 論文貢獻 4
1-4 論文架構 5
第二章文獻探討 6
2-1 不同製程造成的瑕疵樣態 6
2-2 基於密度聚類演算法DBSCAN 7
2-3 混合瑕疵樣態與離群值去除 8
2-4 卷積神經網路(Convolutional Neural Networks，CNN） 9
2-5 無樣態優先的兩階段識別系統(None-First Two-Stage Model) 10
2-6 混合瑕疵樣態識別 11
2-7 基於OPTICS分群的瑕疵樣態識別法 11
2-8 提高深度卷基神經網路訓練的資料增強與降雜訊方法 12
2-9 隨機性特徵參數B-score用於晶圓圖分割分析[20] 13
第三章研究方法 15
3-1 系統流程圖 15
3-2 晶圓圖分群算法 16
3-3 圓形切割特徵分析 19
3-3-1 區域密度(Density of Bad Die, DBD) 19
3-3-2 邊緣壞晶粒比例(Edge BDP) 20
3-3-3 歸一化權重壞晶粒比(NWBDP) 21
3-3-4 壞晶粒重點分布(BD Focus) 23
3-4 群聚濾波器 24
第四章實驗結果與分析 27
4-1 濾波器識別結果 27
4-1-1 平均濾波器 27
4-1-2 四種濾波器識別結果 29
4-1-3 濾波後精度上升與None下降之討論 30
4-2 策略性選擇濾波器的識別結果 31
4-2-1 NFTS第一階段識別討論 33
4-3 識別結果比較 35
4-3-1 改善運算速度 37
第五章結論 40
第六章參考文獻 41

參考文獻

[1] M. J. Wu, J. S. R. Jang, and J. L. Chen, “Wafer map failure pattern recognition and similarity ranking for large-scale data sets,” IEEE Trans. Semicond. Manuf., vol. 28, no. 1, pp. 1-12, Feb. 2015, doi: 10.1109/TSM.2014.2364237.
[2] F. Adly, P. D. Yoo, S. Muhaidat, Y. Al-Hammadi, U. Lee, and M. Ismail, “Randomized general regression network for identification of defect patterns in semiconductor wafer maps,” IEEE Trans. Semicond. Manuf., vol. 28, no. 2, pp. 145-152, May 2015, doi: 10.1109/TSM.2015.2405252.
[3] F. Adly , O. Alhussein, P. D. Yoo, Y. Al-Hammadi, K. Taha, S. Muhaidat, Y.-S. Jeong, U. Lee, and M. Ismail, “Simplified subspaced regression network for identification of defect patterns in semiconductor wafer maps,” IEEE Trans. Industrial Informatics, vol. 11, no. 6, pp. 1267-1276, Dec. 2015, doi: 10.1109/TII.2015.2481719.
[4] M. Piao, C. H. Jin, J. Y. Lee, and J.-Y. Byun, “Decision tree ensemble-based wafer map failure pattern recognition based on radon transform-based features,” IEEE Trans. Semicond. Manuf., vol. 31, no. 2, pp. 250-257, May 2018, doi: 10.1109/TSM.2018.2806931.
[5] G. Tello, O. Y. Al-Jarrah, P. D. Yoo, Y. Al-Hammadi, S. Muhaidat, and U. Lee, “Deep-structured machine learning model for the recognition of mixed-defect patterns in semiconductor fabrication processes,” IEEE Trans. Semicond. Manuf., vol. 31, no. 2, pp. 315-322, May 2018, doi: 10.1109/TSM.2018.2825482.
[6] K. Kyeong and H. Kim, “Classification of mixed-type defect patterns in wafer bin maps using convolutional neural networks,” IEEE Trans. Semicond. Manuf., vol. 31, no. 3, pp. 395-402, Aug. 2018, doi: 10.1109/TSM.2018.2841416.
[7] J. Yu and X. Lu, "Wafer Map Defect Detection and Recognition Using Joint Local and Nonlocal Linear Discriminant Analysis," IEEE Transactions on Semiconductor Manufacturing, vol. 29, no. 1, pp. 33-43, Feb. 2016, doi: 10.1109/TSM.2015.2497264.
[8] Takeshi Nakazawa, and Deepak V. Kulkarni, “Wafer Map Defect Pattern Classiﬁcation and Image Retrieval Using Convolutional Neural Network,” IEEE Transaction on Semiconductor Manufacturing, Vol 31, No. 2, pp. 309-314, May. 2018.
[9] Mill-Jer Wang, Yen-Shung Chang, J.E. Chen, Yung-Yuan Chen, and Shaw-Cherng Shyu, “Yield Improvement by Test Error Cancellation”, in Proc. Asian Test Symposium (ATS′96), pp.258-260, Nov. 1996.
[10] Martin Ester, Hans-Peter Kriegel, Jörg Sander, and Xiaowei Xu, “A density-based algorithm for discovering clusters in large spatial databases with noise,” in Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (KDD′96), pp. 226–231, 1996.

[11] Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, Nov. 1998, doi: 10.1109/5.726791.
[12] Cheng Hao Jin , Hyuk Jun Na , Minghao Piao , Gouchol Pok, and Keun Ho Ryu, "A Novel DBSCAN-Based Defect Pattern Detection and Classification Framework for Wafer Bin Map", IEEE Transactions on Semiconductor Manufacturing, Vol. 32, No. 3, pp. 286 - 292, Aug. 2019.
[13] Jia-Xiu Zhang, "NFTS: A None-First Two-Stage Model for Wafer Map Defect Classification," master′s thesis, National Central University, 2021.
[14] Po-Lin Huang, " Improving None-First Two-Stage Method with Feature Analysis for Wafer Map Defect Classification," master′s thesis, Chung Yuan Christian University, 2022.
[15] A. Rosenfeld and J. Pfaltz, “Sequential operations in digital picture processing,” J. ACM, vol. 13, no. 4, pp. 471–494, 1966.
[16] Hsieh, Liam Y., and Tsung-Ju Hsieh, "A Throughput Management System for Semiconductor Wafer Fabrication Facilities: Design, Systems and Implementation" Processes 2018, Vol. 6, Issue 2, Article No. 16.
[17] C.-W. Chang, T.-M. Chao, J.-T. Horng, C.-F. Lu, and R.-H. Yeh, ‘‘Development pattern recognition model for the classification of circuit probe wafer maps on semiconductors,’’ IEEE Trans. Compon., Packag., Manuf. Technol., vol. 2, no. 12, pp. 2089–2097, Dec. 2012.
[18] Mengying Fan, Qin Wang and B. van der Waal, "Wafer defect patterns recognition based on OPTICS and multi-label classification," in Proc. 2016 IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC), Xi′an, China, 2016, pp. 912-915, doi: 10.1109/IMCEC.2016.7867343.
[19] T. Ishida, I. Nitta, D. Fukuda and Y. Kanazawa, "Deep Learning-Based Wafer-Map Failure Pattern Recognition Framework," in Proc. 20th International Symposium on Quality Electronic Design, Santa Clara, CA, USA, 2019, pp. 291-297, doi: 10.1109/ISQED.2019.8697407.
[20] 呂東穎,“Application of Wafer Map Partition Analysis to Enhance the Salient Pattern Identification Analysis,”master′s thesis, National Central University, 2019

指導教授

陳聿廣梁新聰(Yu-Guang Chen Hsing-Chung Liang)

審核日期

2023-12-19

推文