یک مدل هوش مصنوعی برای ساخت شاخص سلامت چرخ‌دنده‌ها

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشکده مهندسی مکانیک، دانشگاه صنعتی اصفهان، اصفهان، ایران

2 دانشکده مهندسی مکانیک، دانشگاه صنعتی اصفهان، اصفهان، ایران.

3 دانشکده فنی و مهندسی (گروه برق)، دانشگاه آزاد اسلامی واحد مبارکه، اصفهان

چکیده

هدف از پایش وضعیت افزارگان مکانیکی، رصد لحظه‌ای افزار به‌منظور پیش‌بینی رخ داد واماندگی است. اولین گام برای پایش وضعیت یک افزار ساخت شاخص سلامت برای آن است. برای ایجاد شاخص سلامت ابتدا بایست داده‌های عملکردی همچون داده‌های ارتعاشی از افزار در طی مدت بهره‌برداری گردآوری شوند و سپس ویژگی‌های معنی‌دار از داده‌ استخراج شوند. در این مطالعه یک مدل هوش مصنوعی (خود رمزگذار همگشتی[i]) برای استخراج ویژگی از داده‌های ارتعاشی معرفی شده که صرفاً نیازمند داده‌های وضعیت سالم افزار برای آموزش است. در این مدل، داده‌های ارتعاشی افزار در طی مدت بهره‌برداری در وضعیت سالم افزار به‌صورت برخط جمع‌آوری می‌شوند تا پایگاه داده در وضعیت سالم ایجاد شود. پس از تشخیص شروع وضعیت خرابی افزار، افزودن داده به پایگاه داده وضعیت سالم متوقف شده و مدل خود رمزگذار همگشتی توسط پایگاه داده آموزش می‌بیند. درنهایت در طی مرحله رشد خرابی‌ها، شاخص سلامت توسط تفاوت داده‌های ارتعاشی افزار در وضعیت خراب با پایگاه داده در وضعیت سالم ساخته می‌شود. عملکرد مدل پیشنهادی توسط داده‌های ارتعاشی چرخ‌دنده ارزیابی شده است و نتایج نشان‌دهنده عملکرد قابل‌قبول این روش در ساخت شاخص سلامت برای چرخ‌دنده‌ها است.
 
[i]. Convolutional Autoencoder


 


 

 
 

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

An Artificial Intelligence Model for the Construction of a Health Indicator for Gears

نویسندگان [English]

  • mohammadreza kaji 1
  • Jamshid Parvizian 2
  • mohammad Silani 1
  • Sayed Hossein Mirlohi 3
1 Department of Mechanical Engineering, Isfahan University of Technology, Isfahan, Iran
2 Department of Mechanical Engineering, Isfahan University of Technology. Iran
3 Department of Electrical Engineering, Mobarake Branch, Islamic Azad University, Isfahan, Iran
چکیده [English]

The purpose of condition monitoring is to monitor the conditions of an asset in order to predict its failure. The first step in implementing condition monitoring is to establish a health indicator. To construct the health indicator, operational data such as vibrational data should be collected from the asset during its operation, followed by extracting meaningful features from the data.This study introduces an artificial intelligence model (convolutional autoencoder) for feature extraction from the vibration data, which only requires healthy status data for training. For this purpose, the vibration data of the asset is gathered online during operation in a healthy state to establish the healthy dataset. Then, the deep learning model is trained by the healthy dataset. Finally, after the failure stage is detected, the health indicator is established by measuring the differences in the vibrations of the healthy and the failure conditions. The performance of the proposed model is evaluated by vibrational data of a gearbox. The health indicator exhibits a monotonically increasing degradation trend and has good performance in terms of detecting incipient faults.

کلیدواژه‌ها [English]

  • Condition monitoring
  • artificial intelligence
  • health indicator
  • remaining useful life
[1] Wang, Dong, Kwok-Leung Tsui, and Qiang Miao, "Prognostics and health management: A review of vibration-based bearing and gear health indicators", Ieee Access, 2017, Vol.6, pp.665-676.
[2] Mobley, R. Keith, “An introduction to predictive maintenance’, Elsevier, 2002.
[3] Gandhare, Balasaheb S., and Milind Akarte, "Maintenance strategy selection", In Ninth AIMS International Conference on Management, Maharashtra, India: AIMS International, 2012, pp. 1330-1336.
[4] B. A. Gandhare, Milind, "Maintenance Strategy Selection," in The 9th AIMS International Conference on Management, 2012.
[5] Duong, Bach Phi, Sheraz Ali Khan, Dongkoo Shon, Kichang Im, Jeongho Park, Dong-Sun Lim, Byungtae Jang, and Jong-Myon Kim, "A reliable health indicator for fault prognosis of bearings", Sensors, 2018, Vol.18, no.11, p.3740.
[6] Liu, Zepeng, and Long Zhang, "A review of failure modes, condition monitoring and fault diagnosis methods for large-scale wind turbine bearings", Measurement, 2020, Vol.149, p.107002.
[7] Lee, Won Gi, Jin Woo Lee, Min Sung Hong, Sung-Ho Nam, YongHo Jeon, and Moon G. Lee, "Failure diagnosis system for a ball-screw by using vibration signals", Shock and Vibration, 2015.
[8] Yen, Gary G., and K-C. Lin, "Wavelet packet feature extraction for vibration monitoring", IEEE transactions on industrial electronics, 2000, Vol.47, no.3, pp.650-667.
[9] Kar, Chinmaya, and A. R. Mohanty, "Vibration and current transient monitoring for gearbox fault detection using multiresolution Fourier transform", Journal of Sound and Vibration, 2008, Vol.311, no.1-2, pp.109-132.
[10] Al-Badour, Fadi, Mehmet Sunar, and Lahouari Cheded, "Vibration analysis of rotating machinery using time–frequency analysis and wavelet techniques", Mechanical Systems and Signal Processing, 2011, Vol.25, no.6, pp.2083-2101.
[11] Elbouchikhi, Elhoussin, Vincent Choqueuse, Yassine Amirat, Mohamed El Hachemi Benbouzid, and Sylvie Turri, "An efficient Hilbert–Huang transform-based bearing faults detection in induction machines", IEEE Transactions on Energy Conversion, 2017, Vol.32, no.2, pp.401-413.
[12] Torrence, Christopher, and Gilbert P. Compo, "A practical guide to wavelet analysis", Bulletin of the American Meteorological society, 1998, Vol.79, no.1, pp.61-78.
[13] Yoo, Youngji, and Jun-Geol Baek, "A novel image feature for the remaining useful lifetime prediction of bearings based on continuous wavelet transform and convolutional neural network", Applied Sciences, 2018, Vol.8, no.7, p.1102.
[14] Abed, Wathiq, Sanjay Sharma, Robert Sutton, and Amit Motwani, "A robust bearing fault detection and diagnosis technique for brushless DC motors under non-stationary operating conditions", Journal of Control, Automation and Electrical Systems, 2015, Vol.26, pp.241-254.
[15] He, Jun, Ming Ouyang, Chen Yong, Danfeng Chen, Jing Guo, and Yan Zhou, "A novel intelligent fault diagnosis method for rolling bearing based on integrated weight strategy features learning", Sensors, 2020, Vol.20, no.6, p.1774.
[16] Yin, Hang, Zhongzhi Li, Jiankai Zuo, Hedan Liu, Kang Yang, and Fei Li, "Wasserstein generative adversarial network and convolutional neural network (WG-CNN) for bearing fault diagnosis", Mathematical Problems in Engineering, 2020, pp.1-16.
[17] Booyse, Wihan, Daniel N. Wilke, and Stephan Heyns, "Deep digital twins for detection, diagnostics and prognostics", Mechanical Systems and Signal Processing, 2020, Vo.140, p.106612.
[18] Carroll, James, Sofia Koukoura, Alasdair McDonald, Anastasis Charalambous, Stephan Weiss, and Stephen McArthur, "Wind turbine gearbox failure and remaining useful life prediction using machine learning techniques", Wind Energy, 2019, Vol.22, no.3, pp.360-375.
[19] Fan, Xianfeng, and Ming J. Zuo, "Gearbox fault detection using Hilbert and wavelet packet transform", Mechanical systems and Signal processing, 2006, Vol.20, no.4, pp.966-982.
[20] Fan, X. F., H. Lin, and M. J. Zuo, “Gearbox Lifetime Assessment Experiment: Part I, Experiment Design & Operation Guide”, Technical Report. Department of Mechanical Engineering, University of Alberta, Edmonton, Alberta, 2004.
[21] Hua, Cheng, Qing Zhang, Guanghua Xu, Yizhuo Zhang, and Tao Xu, "Performance reliability estimation method based on adaptive failure threshold", Mechanical Systems and Signal Processing, 2013, Vol.36, no.2, pp.505-519.
[22] Ahn, Jeongyoun, Myung Hee Lee, and Jung Ae Lee, "Distance-based outlier detection for high dimension, low sample size data", Journal of Applied Statistics, 2019, Vol.46, no.1. pp.13-29.
[23] Sait, Abdulrahman S., and Yahya I. Sharaf-Eldeen, "A review of gearbox condition monitoring based on vibration analysis techniques diagnostics and prognostics", In Rotating Machinery, Structural Health Monitoring, Shock and Vibration, Volume 5: Proceedings of the 29th IMAC, A Conference on Structural Dynamics, 2011, Springer New York, 2011, pp. 307-324.
[24] Xu, Fan, Zhelin Huang, Fangfang Yang, Dong Wang, and Kwok Leung Tsui, "Constructing a health indicator for roller bearings by using a stacked auto-encoder with an exponential function to eliminate concussion", Applied Soft Computing, 2020, Vol.89 pp.106119.