生活水平的提升使得全球對半導體的需求快速成長，企業為了提升競爭力，致力於改善產能和產品良率，但面板製程十分繁瑣，從原料到成品需經過數百道製程，確保機台正常運作才能避免重工產生的成本。過往的機台維修政策多為運轉到故障才維修和定期保養，但上述維修策略可能影響排定的生產計劃、降低機台利用率。隨著資訊科技的進步，機台感測資料的取得愈趨容易，改變機台維修與提升良率的策略，藉由機台參數的歷史數據適時排定機台維修。透過動態特徵和統計值的萃取，可以從機台感測器產生的時間序列資料中獲取系統衰退、異常現象的行為，並藉由偏最小平方判別分析(partial least squares discrimination analysis, PLS-DA)比較特徵對於機台情況差異的顯著性，針對顯著的特徵建立機台健康指標模型，其模型會將顯著特徵值轉換成與機台狀況相對應的健康指標，並偵測機台狀態異常造成的健康指標超出門檻值之情況。此系統化的機台預測性維修策略，以某半導體廠之機台資料為實證案例，觀察即時製程狀態變異偵測值（status variable identification, SVID）與其建立的模型，提供機台狀況監控的機制、機台狀態異常的預警，進而讓工程師排定維修的時間點，以達到有效的設備健康狀態監控與降低設備停機所造成的損失。

With the increasing demand of consumer electronics, manufacturers dedicate themselves to maintain the competivity, leading production technology and high yield. However, the wafers are produced from ingredients through several complicated porcesses. It is necessary to ensure the status of the fabrication equipments to avoid rework and sudden shutdown. The majority of current equipment maintenance policies are corrective and preventive one. The mentioned policies above interrupt the scheduled production and decrease productivity when the failures or the routine checks happen. With the advancement of manufacturing technology, the sensor data are easily collected from the equipment during process. It triggers the change of the maintenance policy. It is feasible to extract features and statistics to predict the equipment degradation and take the propriate reaction to the certain conditions. The proposed framework applies partial least squares discrimination analysis (PLS-DA) to identify extracted features for equipment status differences. The retained features which is important for the discrimination of status are used for the following modeling. The model transforms features into a corresponding health indicator. It helps users to realize degradation states of the target system. The indicators can be monitored using preset threshold and offer information to make the decision of repair scheduling. We conducted an empirical study in a semiconductor company in Taiwan to validate our research.

目錄 i
表目錄 iii
圖目錄 iv
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 3
1.3 論文結構 3
第二章 文獻回顧 4
2.1 智慧製造 4
2.2 預測性維修 5
2.3 實體模型(Physical model) 6
2.3.1 知識模型(knowledge-based model) 7
2.3.2 資料導向模型(data-driven model) 7
2.3.3 混和模型(combination model) 8
2.4 偏最小平方法(partial least squares) 9
第三章 研究架構 11
3.1 問題定義 13
3.2 資料準備 14
3.3 特徵萃取 15
3.4 特徵選取 18
3.5 機台健康指標模型建立與指標監控 19
3.6 評估與解釋 19
第四章 實證研究 21
4.1 問題定義 21
4.2 資料準備 22
4.3 特徵萃取 23
4.4 特徵選取 24
4.5 機台健康指標模型建立與指標監控 26
4.6 評估與解釋 28
4.6.1 特徵選取演算法之比較 28
4.6.2 分類器演算法之比較 30
第五章 結論 31
5.1 研究貢獻 31
5.2 後續研究方向 32
參考文獻 33

Abdi, H. (2010), "Partial least squares regression and projection on latent structure regression (PLS Regression)," Wiley interdisciplinary reviews: computational statistics, Vol. 2, No. 1, pp. 97-106.
Abe, S. (2010), "Feature Selection and Extraction," in: (eds.), Support Vector Machines for Pattern Classification, Springer London, London, pp. 331-341.
Ahmad, M. W., Mourshed, M., and Rezgui, Y. (2017), "Trees vs Neurons: Comparison between random forest and ANN for high-resolution prediction of building energy consumption," Energy and Buildings, Vol. 147, No., pp. 77-89.
Barker, M. and Rayens, W. (2003), "Partial least squares for discrimination," Journal of Chemometrics: A Journal of the Chemometrics Society, Vol. 17, No. 3, pp. 166-173.
Biswas, S. S., Srivastava, A. K., and Whitehead, D. (2015), "A real-time data-driven algorithm for health diagnosis and prognosis of a circuit breaker trip assembly," IEEE Transactions on Industrial Electronics, Vol. 62, No. 6, pp. 3822-3831.
Bleakie, A. and Djurdjanovic, D. (2013), "Feature extraction, condition monitoring, and fault modeling in semiconductor manufacturing systems," Computers in Industry, Vol. 64, No. 3, pp. 203-213.
Breiman, L. (2001), "Random forests," Machine learning, Vol. 45, No. 1, pp. 5-32.
Carrascal, L. M., Galván, I., and Gordo, O. (2009), "Partial least squares regression as an alternative to current regression methods used in ecology," Oikos, Vol. 118, No. 5, pp. 681-690.

Cassel, C., Hackl, P., and Westlund, A. H. (1999), "Robustness of partial least-squares method for estimating latent variable quality structures," Journal of applied statistics, Vol. 26, No. 4, pp. 435-446.
Chang, P.-L. and Hsu, C.-W. (1998), "The development strategies for Taiwan's semiconductor industry," IEEE Transactions on Engineering Management, Vol. 45, No. 4, pp. 349-356.
Chen, D. and Trivedi, K. S. (2005), "Optimization for condition-based maintenance with semi-Markov decision process," Reliability Engineering & System Safety, Vol. 90, No. 1, pp. 25-29.
Cheng, F., Qu, L., and Qiao, W. (2018), "Fault Prognosis and Remaining Useful Life Prediction of Wind Turbine Gearboxes Using Current Signal Analysis," IEEE Transactions on Sustainable Energy, Vol. 9, No. 1, pp. 157-167.
Chien, C.-F., Hsu, C.-Y., and Chen, P.-N. (2013a), "Semiconductor fault detection and classification for yield enhancement and manufacturing intelligence," Flexible Services and Manufacturing Journal, Vol. 25, No. 3, pp. 367-388.
Chien, C.-F., Hsu, S.-C., and Chen, Y.-J. (2013b), "A system for online detection and classification of wafer bin map defect patterns for manufacturing intelligence," International Journal of Production Research, Vol. 51, No. 8, pp. 2324-2338.
Chouikhi, H., Khatab, A., and Rezg, N. (2014), "A condition-based maintenance policy for a production system under excessive environmental degradation," Journal of Intelligent Manufacturing, Vol. 25, No. 4, pp. 727-737.
Ciabattoni, L., Cimini, G., Ferracuti, F., Grisostomi, M., Ippoliti, G., and Pirro, M. (2015), "Bayes error based feature selection: An electric motors fault detection case study," Proceedings of Industrial Electronics Society, IECON 2015-41st Annual Conference of the IEEE.

Do, P., Voisin, A., Levrat, E., and Iung, B. (2015), "A proactive condition-based maintenance strategy with both perfect and imperfect maintenance actions," Reliability Engineering & System Safety, Vol. 133, No., pp. 22-32.
Field, A. (2009), Discovering statistics using SPSS. Sage publications.
Garcia, M. C., Sanz-Bobi, M. A., and del Pico, J. (2006), "SIMAP: Intelligent System for Predictive Maintenance: Application to the health condition monitoring of a windturbine gearbox," Computers in Industry, Vol. 57, No. 6, pp. 552-568.
Geladi, P. and Kowalski, B. R. (1986), "Partial least-squares regression: a tutorial," Analytica chimica acta, Vol. 185, No., pp. 1-17.
Gondek, C., Hafner, D., and Sampson, O. R. (2016), "Prediction of failures in the air pressure system of scania trucks using a random forest and feature engineering," Proceedings of International Symposium on Intelligent Data Analysis.
Goyal, D. and Pabla, B. (2015), "Condition based maintenance of machine tools—A review," CIRP Journal of Manufacturing Science and Technology, Vol. 10, No., pp. 24-35.
Haq, A. U. and Djurdjanovic, D. (2016), "Virtual metrology concept for predicting defect levels in semiconductor manufacturing," Procedia CIRP, Vol. 57, No., pp. 580-584.
Irajpour, A., Fallahian-Najafabadi, A., Mahbod, M. A., and Karimi, M. (2014), "A framework to determine the effectiveness of maintenance strategies lean thinking approach," Mathematical Problems in Engineering, Vol. 2014, No., pp.
Jung, K., Kulvatunyou, B., Choi, S., and Brundage, M. P. (2016), "An overview of a smart manufacturing system readiness assessment," Proceedings of IFIP International Conference on Advances in Production Management Systems.
Kim, H.-E., Tan, A. C., Mathew, J., and Choi, B.-K. (2012), "Bearing fault prognosis based on health state probability estimation," Expert Systems with Applications, Vol. 39, No. 5, pp. 5200-5213.
Kotsiantis, S., Kanellopoulos, D., and Pintelas, P. (2006), "Data preprocessing for supervised leaning," International Journal of Computer Science, Vol. 1, No. 2, pp. 111-117.
Kuang, T.-H., Yan, Z., and Yao, Y. (2015), "Multivariate fault isolation via variable selection in discriminant analysis," Journal of Process Control, Vol. 35, No., pp. 30-40.
Kumar, S., Vichare, N. M., Dolev, E., and Pecht, M. (2012), "A health indicator method for degradation detection of electronic products," Microelectronics Reliability, Vol. 52, No. 2, pp. 439-445.
Löfsten, H. (2000), "Measuring maintenance performance–in search for a maintenance productivity index," International Journal of Production Economics, Vol. 63, No. 1, pp. 47-58.
Lee, J., Ardakani, H. D., Yang, S., and Bagheri, B. (2015), "Industrial big data analytics and cyber-physical systems for future maintenance & service innovation," Procedia CIRP, Vol. 38, No., pp. 3-7.
Lee, J., Lapira, E., Bagheri, B., and Kao, H.-a. (2013), "Recent advances and trends in predictive manufacturing systems in big data environment," Manufacturing Letters, Vol. 1, No. 1, pp. 38-41.
Liu, X., Li, J., Al-Khalifa, K. N., Hamouda, A. S., Coit, D. W., and Elsayed, E. A. (2013), "Condition-based maintenance for continuously monitored degrading systems with multiple failure modes," IIE Transactions, Vol. 45, No. 4, pp. 422-435.

Lo, Y.-L., Pan, W.-H., Hsu, W.-L., Chien, Y.-C., Chen, J.-Y., Hsu, M.-M., Lou, P.-J., Chen, I.-H., Hildesheim, A., and Chen, C.-J. (2016), "Partial least square discriminant analysis discovered a dietary pattern inversely associated with nasopharyngeal carcinoma risk," PloS one, Vol. 11, No. 6, pp. e0155892.
Marquetti, I., Link, J. V., Lemes, A. L. G., dos Santos Scholz, M. B., Valderrama, P., and Bona, E. (2016), "Partial least square with discriminant analysis and near infrared spectroscopy for evaluation of geographic and genotypic origin of arabica coffee," Computers and Electronics in Agriculture, Vol. 121, No., pp. 313-319.
McIntosh, A., Chau, W., and Protzner, A. (2004), "Spatiotemporal analysis of event-related fMRI data using partial least squares," Neuroimage, Vol. 23, No. 2, pp. 764-775.
Monnin, M., Abichou, B., Voisin, A., and Mozzati, C. (2011), "Fleet historical cases for predictive maintenance," Proceedings of The International Conference Surveillance.
Oussama, A., Elabadi, F., Platikanov, S., Kzaiber, F., and Tauler, R. (2012), "Detection of olive oil adulteration using FT‐IR spectroscopy and PLS with variable importance of projection (VIP) scores," Journal of the American Oil Chemists' Society, Vol. 89, No. 10, pp. 1807-1812.
Patel, R. K. and Giri, V. (2016), "Feature selection and classification of mechanical fault of an induction motor using random forest classifier," Perspectives in Science, Vol. 8, No., pp. 334-337.
Peng, Y., Dong, M., and Zuo, M. J. (2010), "Current status of machine prognostics in condition-based maintenance: a review," The International Journal of Advanced Manufacturing Technology, Vol. 50, No. 1-4, pp. 297-313.

Sait, A. S. and Sharaf-Eldeen, Y. I. (2011), "A Review of Gearbox Condition Monitoring Based on vibration Analysis Techniques Diagnostics and Prognostics," Proceedings of New York, NY.
Sharma, R. K., Kumar, D., and Kumar, P. (2005), "FLM to select suitable maintenance strategy in process industries using MISO model," Journal of Quality in Maintenance Engineering, Vol. 11, No. 4, pp. 359-374.
Shin, J.-H. and Jun, H.-B. (2015), "On condition based maintenance policy," Journal of Computational Design and Engineering, Vol. 2, No. 2, pp. 119-127.
Si, X.-S., Wang, W., Hu, C.-H., and Zhou, D.-H. (2011), "Remaining useful life estimation–a review on the statistical data driven approaches," European journal of operational research, Vol. 213, No. 1, pp. 1-14.
Sánchez-Maroño, N., Alonso-Betanzos, A., and Tombilla-Sanromán, M. (2007), "Filter methods for feature selection–a comparative study," Proceedings of International Conference on Intelligent Data Engineering and Automated Learning.
Strangas, E. G., Aviyente, S., and Zaidi, S. S. H. (2008), "Time–frequency analysis for efficient fault diagnosis and failure prognosis for interior permanent-magnet AC motors," IEEE Transactions on Industrial Electronics, Vol. 55, No. 12, pp. 4191-4199.
Susto, G. A., Schirru, A., Pampuri, S., and Beghi, A. (2012), "A predictive maintenance system based on regularization methods for ion-implantation," Proceedings of 2012 SEMI Advanced Semiconductor Manufacturing Conference.
Tibshirani, R. (1996), "Regression shrinkage and selection via the lasso," Journal of the Royal Statistical Society: Series B (Methodological), Vol. 58, No. 1, pp. 267-288.
Wang, L., Zhang, Z., Long, H., Xu, J., and Liu, R. (2016), "Wind turbine gearbox failure identification with deep neural networks," IEEE Transactions on Industrial Informatics, Vol. 13, No. 3, pp. 1360-1368.
Wold, H. (1966), "Estimation of principal components and related models by iterative least squares," Multivariate analysis, Vol., No., pp. 391-420.
Yan, X., Zhang, C., and Zhang, S. (2003), "Toward databases mining: Pre-processing collected data," Applied Artificial Intelligence, Vol. 17, No. 5-6, pp. 545-561.
Yan, Z. and Yao, Y. (2015), "Variable selection method for fault isolation using least absolute shrinkage and selection operator (LASSO)," Chemometrics and Intelligent Laboratory Systems, Vol. 146, No., pp. 136-146.
Zhang, S., Zhang, C., and Yang, Q. (2003), "Data preparation for data mining," Applied artificial intelligence, Vol. 17, No. 5-6, pp. 375-381.
Zhang, W., Han, J., and Deng, S. (2017), "Heart sound classification based on scaled spectrogram and partial least squares regression," Biomedical Signal Processing and Control, Vol. 32, No., pp. 20-28.
ZhAng, X., KAng, J., Bechhoefer, E., and Zhao, J. (2014), "A new feature extraction method for gear fault diagnosis and prognosis," Eksploatacja i Niezawodność, Vol. 16, No. 2, pp. 295--300.
Zhang, Z., Wang, Y., and Wang, K. (2013), "Fault diagnosis and prognosis using wavelet packet decomposition, Fourier transform and artificial neural network," Journal of Intelligent Manufacturing, Vol. 24, No. 6, pp. 1213-1227.
Zou, J., Chang, Q., Arinez, J., Xiao, G., and Lei, Y. (2017), "Dynamic production system diagnosis and prognosis using model-based data-driven method," Expert Systems with Applications, Vol. 80, No., pp. 200-209.