近年來，先進封裝技術成為半導體產業的研究重點，提升先進封裝製程之品質則成為業界重點專注的研究課題。在快速變遷的市場環境下，面對不斷更新的材料，製造商需要在短時間內穩定製程，生產出高品質的產品來滿足顧客需求，以持續保持市場競爭力。傳統上使用田口方法解決參數設計問題，本研究欲進一步優化參數最佳化問題，以突破田口方法在非線性上的限制。

本研究以台灣某封測廠之晶圓背面研磨製程為例，提出應用類神經網絡與啟發式演算法於晶圓背面研磨製程參數優化。首先，利用類神經網絡模擬出控制因子與輸出反應值的關係建構模型，對於多目標最佳化問題使用望想函數將多個反應值轉換為單一反應值，再分別利用基因演算法、粒子群演算法及模擬退火法，三種啟發式演算法進行參數優化，搜尋最佳因子組合，找出各個控制因子的最佳參數設定值。

本研究成功找出提升晶圓背面研磨製程之最佳參數組合，經過與個案公司取得之原始參數設定及田口方法比較後，結果表明本研究提出之方法明顯取得更好的表現，將可協助案例公司降低生產成本，有效提升製程能力及產品品質。

In recent years, the semiconductor industry has redirected its research towards advanced packaging technologies, placing significant emphasis on enhancing the quality of advanced packaging processes. In a dynamic market environment with frequent material renews, manufacturers must swiftly stabilize the processes and produce high-quality products to meets customer demands and maintain market competitiveness. Traditionally, the Taguchi method has been utilized to address parameter design problems. However, this study seeks to enhance the optimization of parameters to overcome the nonlinear limitations of the Taguchi method.

This study applies an integrated approach using a neural network, heuristic algorithm and the desirability function to optimize the wafer grinding process. Initially, a back-propagation neural network is used to map the nonlinear relationship between the control factors and the corresponding responses based on the experimental data obtained from a semiconductor manufacturing company in Taiwan. Then, the desirability function and three heuristic algorithms are applied to obtain the optimal factor settings.

By comparing the proposed approach with the original approach and the Taguchi method, the results illustrate the superiority and effectiveness of the proposed approach in terms of process capability.

一、英文部分
[1] Agatonovic-Kustrin, S. & Beresford, R. (2000). Basic concepts of artificial neural network (ANN) modeling and its application in pharmaceutical research. Journal of Pharmaceutical and Biomedical Analysis, 22(5), 717-727.
[2] Alavi, A. H. & Gandomi, A. H. (2011). Prediction of principal ground-motion parameters using a hybrid method coupling artificial neural networks and simulated annealing. Computers & Structures, 89(23-24), 2176-2194.
[3] Asadi, E., Da Silva, M. G., Antunes, C. H., Dias, L., & Glicksman, L. (2014). Multi-objective optimization for building retrofit: A model using genetic algorithm and artificial neural network and an application. Energy and Buildings, 81, 444-456.
[4] Chou, C. J. & Chen, L. F. (2012). Combining neural networks and genetic algorithms for optimising the parameter design of the inter-metal dielectric process. International Journal of Production Research, 50(7), 1905-1916.
[5] Derringer, G. & Suich, R. (1980). Simultaneous optimization of several response variables. Journal of Quality Technology, 12(4), 214-219.
[6] Garey, M. R. & Johnson, D. S. (1990). A Guide to the Theory of NP-Completeness. Computers and Intractability, 37-79.
[7] Holland, J. H. (1973). Genetic algorithms and the optimal allocation of trials. SIAM Journal on Computing, 2(2), 88-105.
[8] Hornik, K., Stinchcombe, M., & White, H. (1990). Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks. Neural Networks, 3(5), 551-560.
[9] Kennedy, J. & Eberhart, R. (1995, November). Particle swarm optimization. In Proceedings of ICNN'95-international Conference on Neural Networks, 4, 1942-1948. IEEE.
[10] Kirkpatrick, S., Gelatt Jr, C. D., & Vecchi, M. P. (1983). Optimization by simulated annealing. Science, 220(4598), 671-680.
[11] Liu, B., Wang, R., Zhao, G., Guo, X., Wang, Y., Li, J., & Wang, S. (2020). Prediction of rock mass parameters in the TBM tunnel based on BP neural network integrated simulated annealing algorithm. Tunnelling and Underground Space Technology, 95, 103103.
[12] Ma, H. Y. & Su, C. T. (2010). Applying hierarchical genetic algorithm based neural network and multiple objective evolutionary algorithm to optimize parameter design with dynamic characteristics. 品質學報, 17(4), 311-325.
[13] Metropolis, N., Rosenbluth, A. W., Rosenbluth, M. N., Teller, A. H., & Teller, E. (1953). Equation of state calculations by fast computing machines. The Journal of Chemical Physics, 21(6), 1087-1092.
[14] Shi, Y. (2001, May). Particle swarm optimization: developments, applications and resources. In Proceedings of the 2001 Congress on Evolutionary Computation (IEEE Cat. No. 01TH8546), 1, 81-86. IEEE.
[15] Wang, S. C. (2003). Artificial neural network. Interdisciplinary computing in java programming 2003.
[16] Soepangkat, B. O. P., Norcahyo, R., Effendi, M. K., & Pramujati, B. (2020). Multi-response optimization of carbon fiber reinforced polymer (CFRP) drilling using back propagation neural network-particle swarm optimization (BPNN-PSO). Engineering Science and Technology, an International Journal, 23(3), 700-713.
[17] Su, C. T. & Chang, H. H. (2000). Optimization of parameter design: an intelligent approach using neural network and simulated annealing. International Journal of Systems Science, 31(12), 1543-1549.
[18] Su, C. T. Chen, M. C., & Chan, H. L. (2005). Applying neural network and scatter search to optimize parameter design with dynamic characteristics. Journal of the Operational Research Society, 56, 1132-1140.
[19] Zhang, H. C. & Huang, S. H. (1995). Applications of neural networks in manufacturing: a state-of-the-art survey. The International Journal of Production Research, 33(3), 705-728.

二、中文部分
[1] 林志銘 (2022)，應用六標準差方法提升晶圓背面研磨製程能力，國立清華大學工業工程與工程管理學系研究所，碩士論文。