伴隨著科技日益精進的工業4.0時代下，仍有許多產業採用傳統的維修方式：矯正性維修(corrective maintenance)及預防性維修(preventive maintenance)，高科技產品的生產線倘若發生設備停機便可能導致巨額的生產成本，為追求傳統產業的科技化，新型維護策略顯得相對的重要。而預測性維修(predictive maintenance)是近期迅速竄起的新興維修模式，能夠有效地預測設備故障時間，並事前進行維修使設備無預警失效的風險減至最低。本研究提供一個完善的預測性維修流程，監控分析機械元件的健康狀況，且運用支持向量迴歸(SVR, support vector regression)針對機械元件進行未來狀態的趨勢預測，依照剩餘使用壽命(RUL, remaining useful life)來規劃維修時間點，避免設備無預警停機的狀況發生。本研究在案例分析中，採用滾珠軸承(ball bearing)的監測訊號透過健康指數(HI, health indicator)的建立及預測，得到評估標準平均誤差值(%Er, percent error) -2.36%個位數的良好預測結果，評估分數(Score)更達到0.7024之高分，並且比較文獻Sutrisno et al. (2012)、Hong et al.、(2014) Lei et al.、(2016) 及Guo et al. (2017) 的分析結果。

With the gradually increasing sophistication of technology in the era of Industry 4.0, there are still many industries using traditional maintenance methods: corrective maintenance and preventive maintenance. A shutdown of high-tech production lines could lead to huge amounts of production costs. Traditional industries should apply more sciences and technologies, so the new maintenance strategy appears to be significantly important. However, predictive maintenance is a rapidly emerging maintenance model that can effectively predict the shutdown time of equipment and minimize the risk of equipment failure without warning. This study provides a complete predictive maintenance process to analyze the monitoring health status of mechanical components and predict the future trend by support vector regression. According to the remaining useful life (RUL), to plan the timing of maintenance which could avoid the shutdown of equipment without warning. In this case study, the monitoring signals of ball bearings are used to establish health indicator and predict future trend. The prediction gets a good result that %Er value equals to -2.36% and Score values reaches 0.7024. Finally, the result will be compared with other methods of other literature.

中文文獻
高美卿、劉艷萍、連琨。支持向量回歸機在風電系統槳距角預測中的應用。電子設計工程，18，105-107 (2010)

英文文獻
Fletcher, R., Practical Methods of Optimization, 2nd Edition, John Wiley & Sons, Inc., Chichester (1987).
Guo L., N. P. Li, F. Jia, Y. G. Lei and J. Lin, “A recurrent neural network based health indicator for remaining useful life prediction of bearings,” Neurocomputing, 240, 98-109 (2017).
Hong S., Z. Zhou, E. Zio and K. Hong, “Condition assessment for the performance degradation of bearing based on a combinatorial feature extraction method,” Digital Signal Processing, 27, 159-166 (2014).
Lei Y. G., N. P. Li, S. Gontarz, J. Lin, S. Radkowski and J. Dybala, “A model-based method for remaining useful life prediction of machinery,” IEEE Transactions on Reliability, 65, 1314-1326 (2016).
Li N. P., Y. G. Lei, J. Lin and S. X. Ding, “An improved exponential model for predicting remaining useful life of rolling element bearings,” IEEE Transactions on Industrial Electronics, 62, 7763-7773 (2015).
Mahalanobis, P. C., “On the generalized distance in statistics,” Proceedings of the National Institute of Sciences of India, 2, 49-55, (1936)
Meyer D., E. Dimitriadou, K. Hornik, A. Weingessel, F. Leisch, C. C. Chung and L. C. Chen, “Package e1071,” Repository CRAN (2017).
Qiu H., J. Lee, J. Lin and G. Yu, “Robust performance degradation assessment methods for enhanced rolling element bearing prognostics,” Advanced Engineering Informatics, 17, 127-140 (2003).
Quinlan J. R., “Induction of decision trees,” Machine Learning, 1, 81-106 (1986).
Quinlan J. R., C4.5: Programs for Machine Learning, Morgan Kaufmann Publishers Inc., San Francisco, USA (1993).
Rosenbaum L., A. Dörr, M. R. Bauer, F. M. Boeckler and A. Zel, “Inferring multi-target QSAR models with taxonomy-based multi-task learning,” Journal of Cheminformatics, 5, 1-20 (2013).
Satishkumar R. and V. Sugumaran, “Estimation of remaining useful life of bearings based on support vector regression,” Indian Journal of Science and Technology, 9, 1-7 (2016).
Schölkopf B., A. Smola, K. R. Müller, “Nonlinear component analysis as a kernel eigenvalue problem,” Neural Computation, 10, 1299-1319 (1998).
Schölkopf, B., “Statistical learning and kernel methods,” Technical Report, Microsoft Research, Cambridge (2000).
Smola A. J. and B. Schölkopf, “A tutorial on support vector regression,” Statistics and Computing, 14, 199-222 (2004).
Sutrisno E., H. Oh, A. Vasan and M. Pecht, “Estimation of remaining useful life of ball bearings,” Prognostics and Health Management, 2012 IEEE Conference, Denver, CO, USA (Jun. 18-21 2012).
Vapnik V., The Nature of Statistical Learning Theory, Springer-Verlag New York Inc., New York, USA (1995).
Wang Y., Y. Z. Peng, Y. Y. Zi, X. H. Jin, and K. L. Tsui, “A two-stage data-driven-based prognostic approach for bearing degradation problem,” IEEE Transactions on Industrial Informatics, 11, 924-932 (2016).
Yang F., M. S. Habibullah, T. Zhang, Z. Xu, P. Lim and S. Nadarajan, “Health index-based prognostics for remaining useful life predictions in electrical machines,” IEEE Transactions on Industrial Informatics, 63, 2633-2644 (2016).

網路文獻
IEEE PHM 2012 Prognostic Challenge。網址：http://www.femto-st.fr/en/Research-departments/AS2M/Research-groups/PHM/IEEE-PHM-2012-Data-challenge.php。 上網日期：2017-09-01 (2012) 。
Industrial IoT。網址：https://industrial-iot.com/2017/02/predictive-maintenance-or-predictive-operations/。上網日期：2017-11-01 (2017) 。
MBA智庫百科。網址：http://wiki.mbalib.com/zh-tw/。上網日期：2017-11-01 (2016) 。
Time 時代雜誌。網址：http://time.com/4088793/elevator-max-cloud-thyssenkrupp/。上網日期：2017-11-20 (2015) 。
遠見雜誌。網址：https://www.gvm.com.tw/article.html?id=32232。上網時間：2017-11-20 (2016)。