Practical Options for Adopting Recurrent Neural Network and Its Variants on Remaining Useful Life Prediction

被引:1
|
作者
Youdao Wang [1 ]
Yifan Zhao [1 ]
Sri Addepalli [1 ]
机构
[1] School of Aerospace, Transport and Manufacturing, Cranfield University
基金
英国工程与自然科学研究理事会;
关键词
Remaining useful life prediction; Deep learning; Recurrent neural network; Long short-term memory; Bi-directional long short-term memory; Gated recurrent unit;
D O I
暂无
中图分类号
TP183 [人工神经网络与计算]; TH17 [机械运行与维修];
学科分类号
0802 ; 081104 ; 0812 ; 0835 ; 1405 ;
摘要
The remaining useful life(RUL) of a system is generally predicted by utilising the data collected from the sensors that continuously monitor di erent indicators. Recently, di erent deep learning(DL) techniques have been used for RUL prediction and achieved great success. Because the data is often time-sequential, recurrent neural network(RNN) has attracted significant interests due to its e ciency in dealing with such data. This paper systematically reviews RNN and its variants for RUL prediction, with a specific focus on understanding how di erent components(e.g., types of optimisers and activation functions) or parameters(e.g., sequence length, neuron quantities) a ect their performance. After that, a case study using the well-studied NASA’s C-MAPSS dataset is presented to quantitatively evaluate the influence of various state-of-the-art RNN structures on the RUL prediction performance. The result suggests that the variant methods usually perform better than the original RNN, and among which, Bi-directional Long Short-Term Memory generally has the best performance in terms of stability, precision and accuracy. Certain model structures may fail to produce valid RUL prediction result due to the gradient vanishing or gradient exploring problem if the parameters are not chosen appropriately. It is concluded that parameter tuning is a crucial step to achieve optimal prediction performance.
引用
收藏
页码:45 / 64
页数:20
相关论文
共 50 条
  • [31] RGCNU: Recurrent Graph Convolutional Network With Uncertainty Estimation for Remaining Useful Life Prediction
    Qiwu Zhu
    Qingyu Xiong
    Zhengyi Yang
    Yang Yu
    IEEE/CAAJournalofAutomaticaSinica, 2023, 10 (07) : 1640 - 1642
  • [32] Dilated Convolutional Recurrent Deep Network with Transfer Learning for Remaining Useful Life Prediction
    Lee, Jing Yang
    Das, Ankit K.
    Hussain, Shaista
    Feng, Yang
    TRENDS AND APPLICATIONS IN KNOWLEDGE DISCOVERY AND DATA MINING, 2020, 12237 : 153 - 164
  • [33] Particle Filter and Its Variants for Degradation State Estimation and Remaining Useful Life Prediction
    Gu, Hengyangcan
    Mahmoud, Hassan
    Arun, Raj Lourd
    Liu, Jie
    Ma, Xinyi
    2023 PROGNOSTICS AND HEALTH MANAGEMENT CONFERENCE, PHM, 2023, : 256 - 263
  • [34] A Novel Remaining Useful Life Prediction Method for Hydrogen Fuel Cells Based on the Gated Recurrent Unit Neural Network
    Long, Bing
    Wu, Kunping
    Li, Pengcheng
    Li, Meng
    APPLIED SCIENCES-BASEL, 2022, 12 (01):
  • [35] An application of systemic prediction evaluation parameters for neural network remaining useful life prediction models
    Qiao, Li
    Shi, Junyou
    An, Weiran
    2015 IEEE CONFERENCE ON PROGNOSTICS AND HEALTH MANAGEMENT (PHM), 2015,
  • [36] Gated Recurrent Unit Networks for Remaining Useful Life Prediction
    Li, Li
    Zhao, Zhen
    Zhao, Xiaoxiao
    Lin, Kuo-Yi
    IFAC PAPERSONLINE, 2020, 53 (02): : 10498 - 10504
  • [37] An Autoencoder Gated Recurrent Unit for Remaining Useful Life Prediction
    Lu, Yi-Wei
    Hsu, Chia-Yu
    Huang, Kuang-Chieh
    PROCESSES, 2020, 8 (09)
  • [38] Autoencoder Quasi-Recurrent Neural Networks for Remaining Useful Life Prediction of Engineering Systems
    Cheng, Yiwei
    Hu, Kui
    Wu, Jun
    Zhu, Haiping
    Shao, Xinyu
    IEEE-ASME TRANSACTIONS ON MECHATRONICS, 2022, 27 (02) : 1081 - 1092
  • [39] A novel spatio-temporal hybrid neural network for remaining useful life prediction
    Wang, Tao
    Tang, Xianghong
    Lu, Jianguang
    Liu, Fangjie
    JOURNAL OF SUPERCOMPUTING, 2023, 79 (17): : 19095 - 19117
  • [40] Joint autoencoder-regressor deep neural network for remaining useful life prediction
    Ince, Kuersat
    Genc, Yakup
    ENGINEERING SCIENCE AND TECHNOLOGY-AN INTERNATIONAL JOURNAL-JESTECH, 2023, 41