為滿足近年來電子產品的快速發展，勢必要縮短電子封裝研發量產時間。本篇論文著重於電子封裝的可靠度分析與機器學習之結合應用。目前，電子封裝的壽命分析法有兩種，第一個為業界常用的溫度循環負載測試(Thermal cyclic test，TCT)，第二個是使用有限單元法之模擬分析。
雖然模擬已證實可取代可靠度真實實驗來減少冗長的電子封裝測試時間，然而每當有新數據產生，這兩種電子封裝可靠度測試方法必須藉由繁複的流程來重新建立一組新的模型或實驗。若我們能夠結合過去大量的經驗資料建立一組數據庫搭配機器學習，如此便可省去那些繁雜的製作程序避免實驗與模擬誤差，讓電腦自行做出快速又精準的預測，同時又可減少對可靠度測試及硬體所需的昂貴費用，可謂經濟又實惠。
本研究目的是希望藉由可靠的數據庫搭配機器學習快速預測封裝壽命來追求整體最大效益。本研究結合機器學習提供新穎的封裝壽命預測方法，針對晶圓級尺寸封裝(Wafer Level Chip Scale Packaging，WLCSP)的壽命資料庫搭配隨機森林演算法，目的在於短時間內得到準確的封裝壽命預估。由於溫度循環負載測試的資料不易取得，資料量不足以用於機器學習。因此，本研究的訓練數據集是使用可靠的模擬壽命產生足夠資料點以利進行機器學習應用。最終本研究證實此方法確實可行，快速且精準的取得與驗證模擬相同的預估壽命。

For the sake of the rapid growth of electronic products, it is imperative to shorten the development time of electronic packaging. This research applies machine learning to do the reliability analysis of packaging. nowadays, there were two types of reliability analysis methods for electronic packages. The first one was the board level thermal cyclic test (TCT) which is the commonly used in the industry, and the second was the simulation analysis using the finite element method.
Packaging simulation has been proved to replace reliability experiment to reduce time. However, both of simulation and experiment have to re-build a set of new model for the unknown electronic package through the complicated processes. If we combine a large amount of experienced data to build a database with machine learning, then we can pass the complicated procedure to avoid experimental and simulation errors. But also let the computer make fast and accurate predictions.
The purpose of this study is using machine learning to get quickly packaging lifetime prediction. This study applies machine learning to provide a novel packaging lifetime prediction method. The Wafer Level Chip Scale Packaging (WLCSP) lifetime database applied with random forest algorithm to obtain accurate packaging lifetime prediction in a short time. Since the amount of experimental data is not sufficient for machine learning, the training dataset of this study will use reliable FEM model to generate enough data points for machine learning.This study confirms that this method is indeed feasible and achieves the accurately lifetime prediction as the testing data.

目錄
摘要……………………………………………………………………….I
ABSTRACT II
目錄……………………………………………………………………..IV
圖目錄 …………………………………………………………………..VI
表目錄……………………………………………………………….. IX
第一章 緒論 1
1.1 研究背景 1
1.2研究動機 1
1.3 文獻回顧 2
1.4 研究目標 7
第二章 基礎理論 9
2.1 錫球外型預估 9
2.2 有限元素法基礎理論 11
2.2.1 線彈性有限元素法理論 12
2.2.2 材料非線性理論 16
2.2.3 數值方法及收斂準則 17
2.3 硬化法則 19
2.3.1 等向硬化法則 20
2.3.2 動態硬化法則 20
2.4 CHABOCHE 模型 21
2.5 封裝結構可靠度預測方法 23
2.5.1 Coffin-Manson應變法 24
2.5.2 Darveaux能量密度法 24
2.5.3 修正型能量密度法(Max’s model) 25
2.6 機器學習 25
2.6.1 數據預處理(Preprocessing) 27
2.6.2監督式學習 29
2.6.3人工類神經網絡(Artificial Neural Network，ANN) 36
2.6.4 CART決策樹 (Classification and regression tree) 37
2.6.5 隨機森林演算法 ( Random Forest Algorithm ) 42
2.7 小結 44
第三章 WLCSP有限元素模型及機器學習 46
3.1 WLCSP有限元素模型建立 46
3.2 材料參數設定 54
3.3 基本假設及邊界條件設定 55
3.4 溫度負載設定 55
3.5 網格尺寸控制方法 56
第四章 機器學習預測封裝壽命 59
4.1 機器學習資料庫 59
4.2 隨機森林回歸模型與預測效能 61
4.3 隨機森林特性評估 67
第五章 結論與未來研究建議 79
第六章 參考文獻 82
附錄 程式碼 88

[1] M.H. Hsu and K.N. Chiang, “Investigation of thermal cycling maximum temperature effect on fatigue life of WLCSP”, International Conference on Electronics Packaging, Sapporo, Japan, April 20-22, 2016.
[2] C.Y. Tsou, T.N. Chang, K.C. Wu, P.L. Wu and K.N. Chiang, “Reliability assessment using modified energy based model for WLCSP solder joints”, International Conference on Electronics Packaging, Yamagata, Japan, April 19-22, 2017.
[3]張天寧，“具玻璃基板扇出型晶圓級封裝結構參數分析及可靠度評估”，碩士論文，國立清華大學動力機械工程學系，2017.
[4] 鄒承諺，以修正型能量密度法評估晶圓級封裝之可靠度，碩士論文，國立清華大學動力機械工程學系，2017.
[5]吳配綸，以不同理論進行墊子封裝熱應力模擬之適用有限單元尺寸，碩士論文，國立清華大學動力機械工程學系， 2017.
[6] R. Darveaux, “Effect of simulation methodology on solder joint crack growth correlation and fatigue life prediction”, Journal of Electronic Packaging, vol. 124, pp. 147-154, 2002.
[7] 吳凱強，先進封裝錫球接點於不同溫度循環負載速率下之可靠度評估，國立清華大學動力機械工程學系，博士論文，2016.
[8] R. Tummala, Fundamentals of Microsystems Packaging, 1st ed, McGraw Hill Professional, 2001.
[9] C.M. Liu, C.C. Lee and K.N. Chiang, “Enhancing the reliability of wafer level packaging by using solder joints layout design”, IEEE Transactions on Component and Packaging Technologies, vol. 29, pp. 877-885, 2006.
[10] T.T. Mattila, H. Xu, O. Ratia and M. Paulasto-Krockel, “Effects of thermal cycling parameters on lifetimes and failure mechanism of solder interconnections”, 60th Electronic Components and Technology Conference, Las Vegas, NV, USA, June 1-4, 2010.
[11] Y.S. Chan and S.W.R. Lee, “Detailed investigation on the creep damage accumulation of lead-free solder joints under accelerated temperature cycling”, Thermal, Mechanical & Multi-Physics Simulation, and Experiments in Microelectronics and Microsystems, Bordeaux, France, April 26-28, 2010.
[12] W. N. Findley, "Creep and relaxation of nonlinear viscoelastic materials with an introduction to linear viscoelasticity." Journal of Applied Mechanics, vol. 44, no. 364, 1977.
[13] J. Lau, W. Danksher and P. Vianco, “Acceleration models, constitutive equations, and reliability of lead-free solders and joints”, Electronic Components and Technology Conference, New Orleans, LA, USA, May 27-30, 2003.
[14] K. N. Chiang and W. L. Chen, “Electronic Packaging Reflow Shape Prediction for the Solder Mask Defined Ball Grid Array,” Transactions on Journal of Electronic Packaging, vol. 120, no. 2, pp,175-178, 1998
[15] M. C. Hsieh and S.L. Tzeng, "Solder Joint Fatigue Life Prediction in Large Size and Low Cost Wafer-Level Chip Scale Packages," International Conference on Electronic Packaging Technology(ICEPT) , pp. 496-501, Chengdu, China , Aug. 12-15 2014.
[16] K.N. Chiang, H.C. Cheng and W.H. Chen, “Large-scaled 3-D area array electronic packaging analysis”, Journal of Computer Modeling and Simulation in Engineering, vol. 4, no.1, pp.4-11, 1999.
[17] E. Osuna, R. Freund and F. Girosit, “Training support vector machines: an application to face detection”, Computer Vision and Pattern Recognition, San Juan, PR, USA, June 17-19, 1997.
[18] Y. Amit and D. Geman, “Shape quantization and recognition with randomized trees”, Neural Computation, vol. 9, pp. 1545-1588, 1997.
[19] Κ. Fukushima, “Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position”, Competition and cooperation in neural nets, Springer, Berlin, Heidelberg, pp. 267-285, Feb. 15-19, 1982.
[20] K. Fukushima and N. Wake, “Handwritten alphanumeric character recognition by the neocognitron”, IEEE Transactions Neural Networks, vol. 2, no. 3, pp. 355-365, 1991
[21] X. H. Li, “Using Random Forest for Classification and Regression”, Chinese Journal of Applied Entomology, vol. 50, no. 4, pp. 1190-1197, 2013.
[22] L. Breiman, Classification and Regression Trees, 1st ed, Routledge, 2017.
[23] T. K. Ho, “Random Decision Forests”, IEEE transactions on pattern analysis and machine intelligence, vol. 20, no. 8, pp. 832-844, 1998.
[24] D. R. Cutler, T. C. Edwards, K. H. Beard, A. Cutler, K. T. Hess, J. Gibson, and J. J. Lawler. “Random forests for classification in ecology ”, Ecology, vol. 88, no.11, pp. 2783-2792, 2007.
[25] Grömping, Ulrike. “Variable importance assessment in regression: linear regression versus random forest”, The American Statistician, vol. 63, no. 4, pp. 308-319, 2009.
[26] X. Li, and Y. Wang. “Applying various algorithms for species distribution modelling”, Integrative Zoology, vol. 8, no. 2, pp. 124-135, 2013.
[27] 李航,“統計學方法”, 初版, 北京: 清華大學出版社, 2012.
[28] 周志華,“機器學習”, 初版, 北京: 清華大學出版社, 2016.
[29] W. D. Van Driel, G. Q Zhang, J. H. Janssen, and L. J. Ernst, “Response surface modeling for nonlinear packaging stresses”, Journal of Electronic Packaging, vol. 125, no. 4, pp. 490-497, 2003.
[30] A. Matsunaga, and J. A. Fortes, “On the use of machine learning to predict the time and resources consumed by applications”, Proceedings of the 2010 10th IEEE/ACM International Conference on Cluster, Cloud and Grid Computing, pp. 495-504.
[31] W. S. McCulloch, and W. Pitts, A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, vol. 5, no. 4, pp. 115-133, 1943.
[32] L. Breiman. “Random forests”, Machine learning, vol. 45, no. 1, pp. 5-32, 2001.
[33] S. S. Dong, Z. X. Huang. “A brief theoretical overview of random forests”, Journal of Integration Technology, vol. 2, no. 1, pp. 1-7, 2013.
[34] L. Breiman, “Bagging predictors”, Machine learning, vol. 24, no. 2, pp. 123-140, 1996.
[35] F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, and J. Vanderplas, “Scikit-learn: Machine learning in Python”, Journal of machine learning research, vol. 12, pp. 2825-2830, 2011.
[36] K. N. Chiang, H. C. Cheng, and W. H. Chen, “Large-Scaled 3-D Area Array Electronic Packaging Analysis”, Journal of Computer Modeling and Simulation in Engineering, vol. 4, no.1, pp.4-11, 1999.
[37] L. Anand, “Constitutive Equations for Hot-Working of Metals,” International Journal of Plasticity, vol. 1, pp. 213-231, 1985.
[38] M. C. Hsieh and S.L. Tzeng. “Solder Joint Fatigue Life Prediction in Large Size and Low Cost Wafer-Level Chip Scale Packages”, International Conference on Electronic Packaging Technology (ICEPT), pp. 496-501, Chengdu, China , Aug. 12-15, 2014.
[39] M. C. Hsieh, “Modeling Correlation for Solder Joint Fatigue Life Estimation in Wafer-Level Chip Scale Packages”, International Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), pp. 65–68, Taipei, Taiwan, Oct. 21-23 2015.
[40] T. N. Chang, C. Y. Tsou, B. H. Wang, and K. N. Chiang. “Novel wafer level packaging for large die size device”, International Conference on IEEE, Electronics Packaging (ICEP), pp. 292-296, Yamagata, Japan, Apr. 19–22, 2017.
[41] J. J. Hopfield, “Neural Networks and Physical Systems with Emergent Collective Computational Abilities”, Proceedings of the national academy of sciences, vol. 79, no. 8, pp. 2554-2558, 1982.
[42] J. Kellie, and Ryan V. Kimes. “Empirical characterization of random forest variable importance measures”, Computational Statistics & Data Analysis, vol. 52, no. 4, pp. 2249-2260, 2008.
[43] 劉立民, 吳建華,“Python機器學習”, 初版, 新北市: 博碩文化股份有限公司, 2017.
[44] M. B. Christopher, Pattern Recognition and Machine Learning, 2nd ed., New York: Springer, 2006.
[45] W. S. McCulloch, and W. Pitts, “A Logical Calculus of the Ideas Immanent in Nervous Activity”, The Bulletin Of Mathematical Biophysics, vol. 5, no. 4, pp. 115-133, 1943.
[46] 曾憲雄,“人工智慧與專家系統理論實務應用”, 初版, 台北市: 旗標出版股份有限公司, 2005.
[47] R. Courant, “Variational methods for the solution of problems of equilibrium and vibrations”, Bulletin of the American Mathematical Society, vol. 49, no. 1, pp. 1-24, 1943.
[48] L. F. Coffin, “A Study of the Effects of Cyclic Thermal Stress on a Ductile Metal,” ASME Journal of Electronic Packaging, vol. 76, pp. 931-950, 1954.
[49] Y. J. Xu, L. Q. Wang, F. S. Wu, W. S. Xia, H. Liu, “Effect of Interface Structure on Fatigue Life under Thermal Cycle with SAC305 Solder Joints,” IEEE Electronic Packaging Technology International Conference, pp, 959-964, Dalian, China, Aug. 11-14, 2013
[50] R. Darveaux, “Effect of Simulation Methodology on Solder Joint Crack Growth Correlation and Fatigue Life Prediction,” Journal of Electronic Packaging, vol. 124, pp. 147-154, 2002.
[51] M. J. Pfeifer, “Solder Bump Size and Shape Modeling and Experimental Validation,” IEEE Transactions on Components, Packaging and Manufacturing Technology – Part B, vol. 20, no.4, pp. 452-457,1997.
[52] S. M. Heinrich, M. Schaefer, S. A. Schroeder and P. S. Lee, “Prediction of Solder Joint Geometries in Array-Type Interconnects”, ASME Journal of Electronic Packaging, vol. 118, pp. 114-121,1996
[53] K. A. Brakke, “The Surface Evolver”, Experimental Mathematics, vol.1, no.2, pp. 141-165, 1992.
[54] L. Li and B.H. Yeung, “Wafer Level and Flip Chip Design Through Solder Prediction Models and Validation”, IEEE Transactions on Components and Packaging Technologies, vol. 24, No. 4, pp. 650-654, 2001.
[55] R. D. Cook, D. S. Malkus, M. E. Plesha, and R. J. Witt, Concepts and Applications of Finite Element Analysis 4th, Wiley, 2001.
[56] K. A. Brakke, Surface Evolver Manual, version 2.01 Minneapolis, MN: The Geometry Center, 1996.