隨著科技的進步，生產設備也日趨精密與昂貴，系統的維護也相對的重要。其中預測性維修是對機台進行狀態監測，並在機台發生故障前進行維修，能使機台無預警失效的機率大幅下降。由於現在感測器的普及，精密及昂貴的生產設備都會裝上感測器，隨時監控機台狀態，能有效減少機台忽然停擺的龐大損失，預測維修已成為日後任何設備保養的必備趨勢。故本研究提供一個預測性維修的方法，運用支持向量回歸對物件進行狀態衰退預測，並算出物件剩餘使用壽命，掌握物件健康狀況，並預知可能失效的時間點，以避免系統失效的狀況發生。本研究在案例一中兩個測試零件的剩餘使用壽命如下： (37.10%，80.24%， Sutrisno et al.，2012) ， (5.32%, 2.06%, 本研究) ，可見本研究的結果比較準確，在案例二中使用滾動式預測可以解決長區間預測不準確問題。

The production equipment is becoming more sophisticated and expensive as the progress of science and technology, the maintenance is also more important. Predictive maintenance is to monitor the condition of the machine and repair the machine before it breaks down. It will reduce the probability of the failure of the machine without warning. Due to the popularity of sensors, the precision and expensive production equipment will be installed sensor and monitor state at any time, it can effectively reduce the huge loss of the machine suddenly shut down. Predictive maintenance has become a necessary trend on any of the equipment maintenance. It provides a predictive maintenance method in this study, we use the support vector regression to predict the degradation of object status and calculate the object remaining useful life for predict the possible failure time, it can avoid occurring the system failure. The remaining use life of two test parts in case one is following below: (37.10%, 80.24%, Sutrisno et al.，2012), (5.32%, 2.06%, in this study), the results of this study are more accurate. In case 2, rolling forecast is used to solve the problem of inaccurate prediction of long range.

中文文獻
丁士展、李泓逸、呂亦宸。貨櫃船新船造價影響變數與預測模式之研究 - 整合時間序列分析與灰色理論。航運季刊，24，65-85。(2015)。
張怡筠。CPFR中滾動式預測的關鍵因子探討。國立清華大學資訊系統與應用研究所碩士論文，新竹 (2002)。
辜樹仁。不賣飛機引擎，改賣「飛機時數」。天下雜誌，601，108-111 (2016)。
顧尚芳。生產系統中利用製程不良率評估設備預防維護之研究。中原大學工業工程研究所碩士論文，桃園 (2003)。
英文文獻
Budai-Balke, B., Operations Research Models for Scheduling Railway Infrastructure, PhD thesis, Erasmus University Rotterdam, Netherlands (2009).
Francis, E.G., Systems Reliability and Risk Analysis, Springer Netherlands (1988).
Fletcher, R., Practical Methods of Optimization, 2nd Edition, Wiley (1987).
Gunn, S.R., Support Vector Machines for Classification and Regression, Technical Report, Image speech and Intelligent Systems Research Group, University of Southampton (1998).
Laayouj, N., J. Hicham and M. A. El Hail., “Photovoltaic Module Health Monitoring and Degradation Assessment,” Proceedings of the 2016 3rd Conference on Control and Fault-Tolerant Systems, 7-9 (2016).
Lewis, C.D., Industrial and Business Forecasting Methods, Butterworths, London (1982).
Lin, J., J. Pulido and M. Asplund., “Reliability Analysis for Preventive Maintenance Based on Classical and Bayesian Semi-Parametric Degradation Approaches Using Locomotive Wheel-Sets as a Case Study,” Reliability Engineering and System Safety, 134, 143-156 (2015).
Meyer, D., E. Dimitriadou., K. Hornik., A. Weingessel., F. Leisch., C. Chih Chung. and Lin. Chih Chen., “Package e1071,” Repository CRAN (2017).
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).
Stretch, J., Managing with Rolling Forecasts, Stretch Publishing (2012).
Sutrisno, E., H. Oh., A. Vasan and M. Pecht., “Estimation of Remaining Useful Life of Ball Bearings,” Prognostics and Health Management (PHM), 2012 IEEE Conference, 1-7 (2012).
Vapnik. V., The Nature of Statistical Learning Theory, Springer-Verlag (1995).
網路文獻
Dataconomy。網址：http://dataconomy.com/2015/04/predictive-maintenance-big- data-on-rails/。 上網日期：2017-03-22 (2014) 。
FEMTO-ST Institute。 網址：http://www.femto-st.fr/。 上網日期：2017-04-20 (2012) 。
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2017-04-20 (2012) 。
MBA智庫百科。網址：http://wiki.mbalib.com/zh-tw/。 上網日期：2017-03-22 (2016) 。