簡易檢索 / 詳目顯示

回結果列表
研究生： 洪志宇
Hung, Chi-Yu
論文名稱： 半監督式學習架構於變速下滾動軸承線上損壞檢測
Semi-Supervised Learning Framework for Inline Bearing Diagnosis in Varying Speed
指導教授： 李家岩
Lee, Chia-Yen
洪郁修
Hung, Yu-Hsiu
學位類別： 碩士
Master
系所名稱： 工學院 - 智慧製造國際碩士學位學程
International Master Program on Intelligent Manufacturing
論文出版年： 2022
畢業學年度： 110
語文別： 英文
論文頁數： 95
中文關鍵詞： 故障檢測 、線上診斷 、異常檢測 、一類分類 、半監督式學習 、非高斯背景雜訊 、變轉速操作 、滾動軸承 、濾波器組 、包絡分析 、階次追蹤
外文關鍵詞： Fault Detection, Inline Fault Diagnosis, Novelty Detection, One-class Classification, Semi-Supervised Learning, Non-Gaussian Background Noise, Varying Rotational Speed, Rolling Bearing, Filter Bank, Envelope Analysis, Order Tracking
相關次數： 點閱：243  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 隨著先進製程技術的進步，以及因應高速生產化的需求，機台的故障偵測是值得持續精進的課題。因此相關的研究議題蓬勃發展，除了長年來受到許多專家學者的關注，近年更是受到業界重視並大量導入產線應用中，以期降低因異常所造成的損失。
    本研究針對滾動軸承的故障偵測，並以Randall與Antoni所提出的Fast kurtogram[1, 2]之盲辨識檢測為基礎，提出適合即時自我監控的指標。由於軸承缺陷所激發出的訊號在大量嘈雜雜訊中相對微弱且易受到隨機亂源的影響，甚至在變速運轉工況下更加難以察覺，因此，基於軸承故障產生的訊號特性，本研究提出了一個新穎的指標ORgram，其中OR為Outlier Rate的縮寫。 ORgram的解決方案為: (1) 透過健康狀態資料在每個頻帶個別建模，並應用半監督式學習中的異常檢測方法(一類分類)，判斷測試資料形成最多離群值數量的頻帶。藉由此方法進行頻帶選擇，同時避開隨機亂源帶來的影響；(2) 透過速度曲線自適應性地分段訊號，以處理變速工況所造成提取特徵的困難。由於允許在非高斯雜訊中以及變轉速下進行自我檢測，因而適用於線上診斷。同時，透過包絡分析與階次追蹤等技術，來識別產生損傷的軸承部件，並設計一個機制評價頻帶的選擇結果，以便在軸承損傷擴大前提供更換預警，從而減少意外停機時間。

    With the advancement of advanced process technology and the need for high-speed production in manufacturing, fault detection is a topic worthy of continuous improvement. Therefore, related research topics have not only attracted the attention of many experts and scholars for many years but have also been widely used in production lines by the industry in recent years, in order to reduce the losses caused by abnormalities.
    This research aims to the fault detection of rolling bearing and proposes an indicator based on the Fast kurtogram proposed by Randall and Antoni [1, 2]. The signal excited by the bearing defect is relatively weak in strong background noise, and is more difficult to be detected in varying speed condition. Based on the signal characteristics of bearing defect, this study proposes a novel indicator called ORgram. “OR” is the abbreviation for “Outlier Rate”. The solution proposed by ORgram is as follows: (1) Modeling with healthy bearing data for each frequency band to detect test data with anomaly detection (One-class classification) in semi-supervised learning. And the frequency band with the highest number of outliers will be selected as the most sensitive frequency for diagnosis, which can reduce the interference of random strong noise; (2) The signal is adaptively segmented with the speed profile to deal with the difficulty of feature extraction in varying speed condition. ORgram is suitable for inline diagnosis by allowing bearing self-detection in non-Gaussian background noise and in varying speed condition. In addition, techniques such as envelope analysis and order tracking are used to identify damaged bearing components. And a method for evaluating the result of frequency band selection is constructed for an alarm to reduce unplanned downtime.

    中文摘要 I Abstract II Acknowledgements III Table of Contents III List of Figures IX List of Tables XII Nomenclature XIII Chapter 1. Introduction 1 1.1 Background and Motivation 1 1.1.1 Background 1 1.1.2 Motivation 2 1.2 Problem Description and Research Purposes 6 1.3 Thesis Architecture 7 Chapter 2. Literature Review 8 2.1 Rolling Bearing Life and Bearing Defect 8 2.1.1 Rolling Bearing Life 8 2.1.2 Impulsiveness and Cyclostationarity 13 2.1.3 Rolling Bearing Component and Fault Characteristic Frequency 15 2.2 Blind Signal Separation for Bearing Defect 18 2.2.1 Statistical Indicator 18 2.2.1.1 Spectral kurtosis 21 2.2.1.2 Infogram 23 2.2.1.3 Accugram 28 2.2.2 Filter Bank 29 2.2.2.1 Fast Kurtogram 31 Chapter 3. Methodology 35 3.1 Methodology Flowchart 36 3.1.1 Architecture of ORgram 37 3.2 Filter Bank 38 3.3 Statistical Indicator – Outlier Rate 39 3.3.1 Signal Segmentation and Feature Extraction 39 3.3.2 One-class Classifier 44 3.3.3 Outlier Rate Calculation 46 3.4 Envelope Analysis 48 3.5 Order Tracking 51 3.5.1 Angular Resampling 51 3.5.2 Order 52 3.5.3 The Aim of Order Tracking 53 3.6 Envelope Spectrum Evaluation 54 Chapter 4. Case Study 57 4.1 Signal Simulation 57 4.1.1 Single Impulse Response Simulation 57 4.1.2 Cyclic Pulse Train Simulation in Constant Speed Operation 58 4.1.3 Cyclic Pulse Train Simulation in Varying Speed Operation 59 4.1.4 Model Validation and Comparison 60 4.2 Open Dataset Validation 73 4.2.1 Outer Ring Defect in Increasing Speed 75 4.2.2 Inner Ring Defect in Decreasing Speed 81 4.3 Summary of Case Study 87 Chapter 5. Conclusion and Future Research 89 5.1 Summary and Contribution 89 5.2 Future Work 91 Reference 92 Appendix.1 Nov-local Outlier Factor for Novelty Detection 95

    1. Antoni, J. and R.B. Randall, The spectral kurtosis: application to the vibratory surveillance and diagnostics of rotating machines. Mechanical systems and signal processing, 2006. 20(2): p. 308-331.
    2. Antoni, J., Fast computation of the kurtogram for the detection of transient faults. Mechanical Systems and Signal Processing, 2007. 21(1): p. 108-124.
    3. Bazurto, A.J., E.C. Quispe, and R.C. Mendoza. Causes and failures classification of industrial electric motor. in 2016 IEEE ANDESCON. 2016. IEEE.
    4. Nandi, S., H.A. Toliyat, and X. Li, Condition monitoring and fault diagnosis of electrical motors—A review. IEEE transactions on energy conversion, 2005. 20(4): p. 719-729.
    5. Bindu, S. and V.V. Thomas. Diagnoses of internal faults of three phase squirrel cage induction motor—A review. in 2014 International Conference on Advances in Energy Conversion Technologies (ICAECT). 2014. IEEE.
    6. Bonnett, A.H. and G.C. Soukup, Rotor failures in squirrel cage induction motors. IEEE transactions on industry applications, 1986(6): p. 1165-1173.
    7. Siddique, A., G. Yadava, and B. Singh, A review of stator fault monitoring techniques of induction motors. IEEE transactions on energy conversion, 2005. 20(1): p. 106-114.
    8. Bhowmik, P.S., S. Pradhan, and M. Prakash, Fault diagnostic and monitoring methods of induction motor: a review. International Journal of Applied Control, Electrical and Electronics, 2013. 1: p. 1-18.
    9. Bonnett, A.H. and C. Yung, Increased efficiency versus increased reliability. IEEE Industry Applications Magazine, 2008. 14(1): p. 29-36.
    10. Lebaroud, A. and G. Clerc, Classification of induction machine faults by optimal time–frequency representations. IEEE Transactions on Industrial electronics, 2008. 55(12): p. 4290-4298.
    11. Sawalhi, N. and R.B. Randall, Gear parameter identification in a wind turbine gearbox using vibration signals. Mechanical Systems and Signal Processing, 2014. 42(1-2): p. 368-376.
    12. Hebda-Sobkowicz, J., et al., Informative frequency band selection in the presence of non-Gaussian noise–a novel approach based on the conditional variance statistic with application to bearing fault diagnosis. Mechanical Systems and Signal Processing, 2020. 145: p. 106971.
    13. Katz, S. 5 Ways to Prevent Bearing Failures. Available from: https://www.machinerylubrication.com/Read/28854/prevent-bearing-failures.
    14. Randall, R.B., Vibration-based condition monitoring: industrial, automotive and aerospace applications. 2021: John Wiley & Sons.
    15. Roque, A., et al., An approach to fault diagnosis of rolling bearings. WSEAS transactions on systems and control, 2009. 4(4): p. 188-197.
    16. Sawalhi, N. and R.B. Randall. The application of spectral kurtosis to bearing diagnostics. in Proceedings of ACOUSTICS. 2004.
    17. McDonald, G.L., Q. Zhao, and M.J. Zuo, Maximum correlated Kurtosis deconvolution and application on gear tooth chip fault detection. Mechanical Systems and Signal Processing, 2012. 33: p. 237-255.
    18. Miao, Y., M. Zhao, and J. Lin, Improvement of kurtosis-guided-grams via Gini index for bearing fault feature identification. Measurement Science and Technology, 2017. 28(12): p. 125001.
    19. Hebda-Sobkowicz, J., R. Zimroz, and A. Wyłomańska, Selection of the Informative Frequency Band in a Bearing Fault Diagnosis in the Presence of Non-Gaussian Noise—Comparison of Recently Developed Methods. Applied Sciences, 2020. 10(8): p. 2657.
    20. Liang, K., et al., A novel indicator to improve fast kurtogram for the health monitoring of rolling bearing. IEEE Sensors Journal, 2020. 20(20): p. 12252-12261.
    21. Antoni, J., et al., Cyclostationary modelling of rotating machine vibration signals. Mechanical systems and signal processing, 2004. 18(6): p. 1285-1314.
    22. Borghesani, P., P. Pennacchi, and S. Chatterton, The relationship between kurtosis-and envelope-based indexes for the diagnostic of rolling element bearings. Mechanical Systems and Signal Processing, 2014. 43(1-2): p. 25-43.
    23. Barszcz, T. and A. JabŁoński, A novel method for the optimal band selection for vibration signal demodulation and comparison with the Kurtogram. Mechanical systems and signal processing, 2011. 25(1): p. 431-451.
    24. Antoni, J., The infogram: Entropic evidence of the signature of repetitive transients. Mechanical Systems and Signal Processing, 2016. 74: p. 73-94.
    25. Liu, Z., et al., ACCUGRAM: A novel approach based on classification to frequency band selection for rotating machinery fault diagnosis. ISA transactions, 2019. 95: p. 346-357.
    26. Clausius, R., X. On a modified form of the second fundamental theorem in the mechanical theory of heat. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 1856. 12(77): p. 81-98.
    27. Shannon, C.E., A mathematical theory of communication. The Bell system technical journal, 1948. 27(3): p. 379-423.
    28. Cohen, J.E., Information theory and music. Behavioral Science, 1962. 7(2): p. 137-163.
    29. Vries, C.M.D. and S. Geva. Document Clustering with K-tree. in International Workshop of the Initiative for the Evaluation of XML Retrieval. 2008. Springer.
    30. Antoni, J., Cyclostationarity by examples. Mechanical Systems and Signal Processing, 2009. 23(4): p. 987-1036.
    31. Hirschman, I.I., A note on entropy. American journal of mathematics, 1957. 79(1): p. 152-156.
    32. Wehner, S. and A. Winter, Entropic uncertainty relations—a survey. New Journal of Physics, 2010. 12(2): p. 025009.
    33. Wang, T., et al., A new SKRgram based demodulation technique for planet bearing fault detection. Journal of Sound and Vibration, 2016. 385: p. 330-349.
    34. Xu, Y., et al., Adaptive Kurtogram and its applications in rolling bearing fault diagnosis. Mechanical Systems and Signal Processing, 2019. 130: p. 87-107.
    35. Wan, S. and X. Zhang, Teager energy entropy ratio of wavelet packet transform and its application in bearing fault diagnosis. Entropy, 2018. 20(5): p. 388.
    36. Fyfe, K. and E. Munck, Analysis of computed order tracking. Mechanical systems and signal processing, 1997. 11(2): p. 187-205.
    37. Brandt, A., et al., Main principles and limitations of current order tracking methods. Sound and Vibration, 2005. 39(3): p. 19-22.
    38. Villa, L.F., et al., Angular resampling for vibration analysis in wind turbines under non-linear speed fluctuation. Mechanical Systems and Signal Processing, 2011. 25(6): p. 2157-2168.
    39. What's an Order? 2020/10/07; Available from: https://community.sw.siemens.com/s/article/what-s-an-order.
    40. Leys, C., et al., Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. Journal of experimental social psychology, 2013. 49(4): p. 764-766.
    41. McFadden, P. and J. Smith, Vibration monitoring of rolling element bearings by the high-frequency resonance technique—a review. Tribology international, 1984. 17(1): p. 3-10.
    42. D’Elia, G., et al. Step-by-step algorithm for the simulation of faulted bear-ings in non-stationary conditions. in 27th International Conference on Noise and Vibration Engineering, ISMA 2016 and International Conference on Uncertainty in Structural Dynamics, USD2016. 2016. KU Leuven, Departement Werktuigkunde.
    43. Ho, D. and R. Randall, Optimisation of bearing diagnostic techniques using simulated and actual bearing fault signals. Mechanical systems and signal processing, 2000. 14(5): p. 763-788.
    44. Roberts, J., On the response of a simple oscillator to random impulses. Journal of Sound and Vibration, 1966. 4(1): p. 51-61.
    45. Antoni, J., Cyclic spectral analysis of rolling-element bearing signals: Facts and fictions. Journal of Sound and vibration, 2007. 304(3-5): p. 497-529.
    46. Huang, H. and N. Baddour, Bearing vibration data collected under time-varying rotational speed conditions. Data in brief, 2018. 21: p. 1745-1749.
    47. Abboud, D., et al., Deterministic-random separation in nonstationary regime. Journal of Sound and Vibration, 2016. 362: p. 305-326.
    48. Stander, C. and P. Heyns, Transmission path phase compensation for gear monitoring under fluctuating load conditions. Mechanical systems and signal processing, 2006. 20(7): p. 1511-1522.
    49. Abboud, D., et al., Envelope analysis of rotating machine vibrations in variable speed conditions: A comprehensive treatment. Mechanical Systems and Signal Processing, 2017. 84: p. 200-226.
    50. Mauricio, A., et al., Improved Envelope Spectrum via Feature Optimisation-gram (IESFOgram): A novel tool for rolling element bearing diagnostics under non-stationary operating conditions. Mechanical Systems and Signal Processing, 2020. 144: p. 106891.
    51. Zhao, Z., et al., A weighted multi-scale dictionary learning model and its applications on bearing fault diagnosis. Journal of Sound and Vibration, 2019. 446: p. 429-452.
    52. Peeters, C., P. Guillaume, and J. Helsen. Signal pre-processing using cepstral editing for vibration-based bearing fault detection. in In proc. int. conf. noise vib. eng.(isma). 2016.
    53. Klausen, A., H.V. Khang, and K.G. Robbersmyr, Multi-band identification for enhancing bearing fault detection in variable speed conditions. Mechanical Systems and Signal Processing, 2020. 139: p. 106422.
    54. Borghesani, P., et al., Application of cepstrum pre-whitening for the diagnosis of bearing faults under variable speed conditions. Mechanical Systems and Signal Processing, 2013. 36(2): p. 370-384.

    無法下載圖示 校內：2026-08-01公開
    校外：2026-08-01公開
    電子論文尚未授權公開，紙本請查館藏目錄
    QR CODE