強化卷積神經網路深度學習用於剩餘可用壽命預估

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：91

、訪客IP：3.137.180.122

姓名

郭昶逵(Chang-Kuei, Kuo) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

強化卷積神經網路深度學習用於剩餘可用壽命預估
(Enhancing Convolutional Neural Network Deep Learning for Remaining Useful Life Estimation)

相關論文

★ 以IEEE 802.11為基礎行動隨意無線網路之混合式省電通訊協定	★ 以范諾圖為基礎的對等式網路虛擬環境相鄰節點一致性研究
★ 行動隨意網路可調適及可延展之位置服務協定	★ 同儕式網路虛擬環境高效率互動範圍群播
★ 巨量多人線上遊戲之同儕網路互動範圍語音交談	★ 基於范諾圖之同儕式網路虛擬環境狀態管理
★ 利用多變量分析之多人線上遊戲信任使用者選擇	★ 無位置資訊無線感測網路之覆蓋及連通維持
★ 同儕網路虛擬環境3D串流同儕選擇策略	★ 一個使用802.11與RFID技術的無所不在導覽系統U-Guide之設計與實作
★ 同儕式三維資料串流	★ IM Finder: 透過即時通訊網路線上使用者找尋解答
★ 無位置資訊無線感測網路自走車有向天線導航與協調演算法	★ 多匯點無線感測網路省能及流量分散事件輪廓追蹤
★ 頻寬感知同儕式3D串流	★ 無線感測網路旋轉指向天線定位法

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

近年來，智慧工廠為熱門的研究項目之一。預診斷健康管理系統在智慧工廠應用中扮演相當重要的角色，可位機器或元件提供不同層次的預診斷，如錯誤預測和剩餘使用壽命(Remaining Useful Life，RUL)預估。本篇論文強化卷積神經網路(Convolutional Neural Network, CNN)用於預診斷健康管理的可用壽命預估，意指在機器經過一定時間的運作之後，預估其在無法運作之前所剩餘的可用壽命。透過預估機器可用壽命，可使維護人員能夠提早進行機器的維護或替換，以確保系統的永續性。
本論文提出強化卷積神經網路(Enhanced Convolutional Neural Networks, ECNN)進行機器剩餘可用壽命預估。ECNN與傳統CNN的不同在於，傳統CNN大多數用於圖像處理，而ECNN則應用於時序資料進行剩餘可用壽命預估上。ECNN透過更細緻的資料前處理、更好的適應時機預估(adaptive moment estimation, Adam)優化器、更適合的softplus激活函數以及疊加圖層的方式，將時序資料視為雙色圖像，擷取長時間序列資料的有效特徵，準確預估機器剩餘可用壽命。
本論文在實驗階段以NASA C-MAPSS(Commercial Modular Aero-Propulsion System Simulation)的模擬資料集作為驗證對象，其資料集內容為航太引擎資料之時序資料及運作狀態。我們將驗證結果與其他文獻所提出的方法進行比較，如MLP、SVR、RVR和CNN，可發現ECNN優於其他方法。

摘要(英)

The smart factory becomes a hot research topic recently. Prognostic health management (PHM) plays a critical role in smart factory applications to produce different level of prognostics, such as failure prediction and remaining useful life (RUL) estimation, for machines or components. This thesis enhances the convolutional neural network (CNN) deep learning for RUL estimation in smart factory applications. A CNN is a special type of deep neural networks (DNNs) used in deep learning for analyzing image data for the applications of image recognition and video recognition. It has convolution layers, pooling layers, and fully connected layers. A convolution layer contains many filters to abstract features from input data, and a pooling layer can reduce data dimensionality without losing features. The CNN deep learning has been applied in an earlier study for RUL estimation. This thesis enhances the learning by applying more sophisticated data pre-processing, a better optimizer, namely the adaptive moment estimation (Adam) method, and a proper activation function, namely the softplus function. The enhanced CNN deep learning is applied to NASA C-MAPSS (Commercial Modular Aero-Propulsion System Simulation) data set to estimate the RUL of aero-propulsion engines. Its performance is evaluated by a scoring function that can measure the RUL estimation accuracy. The evaluation results are compared with those of other methods using the multi-layer perceptron (MLP), support vector regression (SVR), relevance vector regression (RVR) and traditional CNN. We find that the enhanced CNN deep learning method is superior to other methods.

關鍵字(中)

★ 智慧工廠
★ 預診斷健康管理系統
★ 剩餘時間壽命
★ 卷積神經網路
★ 深度學習

關鍵字(英)

★ smart factory
★ prognostic health management
★ remaining useful life
★ convolutional neural network
★ deep learning

論文目次

中文摘要 I
ABSTRACT II
誌謝 III
圖目錄 VI
表目錄 VII
一、緒論 1
1.1 研究背景與動機 1
1.2 研究目的與貢獻 1
1.3 論文架構 2
二、背景知識 3
2.1 預診斷技術 3
2.1.1 數據導向方法(Data-Driven Prognostics) 3
2.1.2 模型方法(Model-Based Prognostics) 4
2.1.3 混合方法(Hybrid Approaches) 5
2.2 剩餘可用壽命 5
2.3 類神經網路 5
2-4 深度神經網路 8
2-5 卷積神經網路 9
2.5.1 卷積層(Convolutional Layer) 10
2.5.2 池化層(Pooling Layer) 12
2.5.3 全連接層(Full Connected Layer) 13
2.5.4 激活函數(Activation Function) 14
2.5.5 傳播(Propagation) 16
2.5.6 相關文獻探討 17
2-6 資料正規化 19
三、問題與方法 20
3.1 問題定義 20
3.1.1 資料集 20
3.1.2 Scoring Function 22
3.1.3 剩餘可用壽命目標函數定義 23
3.2 資料前處理 24
3.2.1 運作狀況分群 24
3.2.2 資料正規化方法 25
四、實驗架構與結果 26
4.1 強化卷積神經網路(Enhancing Convolution Neural Network) 26
4.2 架構說明 26
4.2.1 卷積層(Convolutional Layer) 27
4.2.2 池化層(Pooling Layer) 28
4.2.3 訓練最佳化 29
4.3 實驗結果 31
五、結論 36
參考文獻 37

參考文獻

[1] X.-S. Si, W. Wang, C.-H. Hu, D.-H. Zhou, ”Remaining useful life estimation – A review on the statistical data driven approaches”, European Journal of Operational Research, pp. 1-1, 2011
[2] 幣圖誌:
http://www.bituzi.com/2014/11/ann-makes-computer-learn.html
[3] Giduthuri Sateesh Babu, Peilin Zhao, and Xiao-Li Li, “Deep Convolutional Neural Network Based Regression Approach for Estimation of Remaining Useful Life”, S.B. Navathe et al. (Eds.): DASFAA 2016, Part I, LNCS 9642, pp. 214–228, 2016.
[4] Convolutional Neural Network: http://speech.ee.ntu.edu.tw/~tlkagk/courses_ML16.html
[5] How the backpropagation algorithm works:
http://neuralnetworksanddeeplearning.com/chap2.html
[6] Olivier Janssens, Viktor Slavkovikj, Bram Vervisch , Kurt Stockman ,Mia Loccufier, Steven Verstockt , Rik Van de Walle , Sofie Van Hoecke, “Convolutional Neural Network Based Fault Detection for Rotating Machinery”, Journal of Sound and Vibration , pp.331–345, 2016
[7] Saxena, A., Goebel, K., “PHM08 challenge data set. NASA AMES prognostics data repository”, Technical report, Moffett Field, CA, 2008
[8] Pingfeng Wang, Byeng D. Youn, Chao Hu, “A generic probabilistic framework for structural health prognostics and uncertainty management”, Mechanical Systems and Signal Processing 28, 622–637, 2012
[9] Heimes, F.O., “Recurrent neural networks for remaining useful life estimation. In:
International Conference on Prognostics and Health Management”, PHM 2008,
pp. 1–6, October 2008
[10] Rumelhart, D.E., Hinton, G.E., Williams, R.J., “Learning representations by
back-propagating errors.”, In: Anderson, J.A., Rosenfeld, E. (eds.) Neurocomputing:Foundations of Research, pp. 696–699. MIT Press, Cambridge, 1988:
http://dl.acm.org/citation.cfm?id=65669.104451
[11] Chang, C.C., Lin, C.J., “LIBSVM: a library for support vector machines”, ACM Trans. Intell. Syst. Technol. 2, 27:1–27:27, 2011
[12] Tipping, M.E.: The relevance vector machine. In: Solla, S.A., Leen, T.K.,
M¨uller, K.R. (eds.) Advances in Neural Information Processing Systems, vol. 12,
pp. 652–658. MIT Press, Cambridge, 2000
[13] Keras Documentation: https://keras.io/
[14] Sebastian Ruder, “An overview of gradient descent optimization algorithms.”: http://sebastianruder.com/optimizing-gradient-descent/
[15] Zheng, Y., Liu, Q., Chen, E., Ge, Y., Zhao, J.L., “Time series classification using
multi-channels deep convolutional neural networks.”, In: Li, F., Li, G., Hwang, S.,Yao, B., Zhang, Z. (eds.) WAIM 2014. LNCS, vol. 8485, pp. 298–310. Springer,
Heidelberg, 2014
[16] Yang, J.B., Nguyen, M.N., San, P.P., Li, X.L., Krishnaswamy, S., ”Deep convolutional neural networks on multichannel time series for human activity recognition.”, In: Proceedings of the 24th International Conference on Artificial Intelligence, pp. 3995–4001. AAAI Press (2015)
[17] Ramasso, E., Saxena, A., “Review and analysis of algorithmic approaches developed for prognostics on CMAPSS dataset.”, Ann. Conf. Prognostics Health Manag. Soc. 2014, 1–11, 2014

指導教授

江振瑞

審核日期

2017-7-26

推文