计算机科学

计算机科学 ›› 2022, Vol. 49 ›› Issue (11A): 211100285-9.doi: 10.11896/jsjkx.211100285

• 交叉&应用 • 上一篇    下一篇

基于机器学习的剩余使用寿命预测实证研究

王加昌1, 郑代威2, 唐雷1, 郑丹晨1, 刘梦娟2   

  1. 1 中国核动力研究设计院核反应堆系统设计技术重点实验室 成都 610213
    2 电子科技大学网络与数据安全四川省重点实验室 成都 610054
  • 出版日期:2022-11-10 发布日期:2022-11-21
  • 通讯作者: 王加昌(jiachang.wang@163.com)
  • 基金资助:
    国家自然科学基金(61202445);中央高校基本业务费项目(ZYGX2016J096)

Empirical Research on Remaining Useful Life Prediction Based on Machine Learning

WANG Jia-chang1, ZHENG Dai-wei2, TANG Lei1, ZHENG Dan-chen1, LIU Meng-juan2   

  1. 1 Science and Technology on Reactor System Design Technology Laboratory,Nuclear Power Institute of China,Chengdu 610213,China
    2 Network and Data Security Key Laboratory of Sichuan Province,University of Electronic Science and Technology of China,Chengdu 610054,China
  • Online:2022-11-10 Published:2022-11-21
  • About author:WANG Jia-chang,born in 1978,postgraduate,senior engineer.His main research interests include nuclear reactor simulation and data mining.
  • Supported by:
    Nationl Natural Science Foundation of China(61202445) and Fundamental Research Funds for the Central Universities of Ministry of Education of China(ZYGX2016J096).

PDF (PC)

477

摘要： 剩余寿命预测是设备预测性维护的3个核心任务之一。目前最新的研究进展是利用机器学习来建立剩余使用寿命预测模型。论文首先梳理了设备剩余使用寿命预测主要采用的机器学习模型,包括支持向量回归模型、多层感知机模型、卷积神经网络和循环神经网络;然后介绍了3个在剩余使用寿命(Remaining Useful Life,RUL)预测中主要采用的公开数据集,以及两个广泛采用的预测性能评价指标。特色之处是基于NASA提供的涡扇发动机仿真数据集C-MAPSS展示了RUL预测建模的基本步骤和关键技术细节,详细比较了几种代表性预测模型的性能。实验结果显示浅层结构的支持向量回归模型的性能确实显著弱于包含深度神经网络的模型;而在深度神经网络中,卷积神经网络和循环神经网络又显示出了各自在挖掘复杂特征交互以及时序特征交互之间的强大能力。最后展望了剩余寿命预测技术的发展前景并讨论了面临的主要挑战。

关键词: 预测性维护, 剩余使用寿命预测, 机器学习, 支持向量回归, 多层感知机, 卷积神经网络, 循环神经网络

Abstract: Remaining useful life(RUL) prediction is one essential task of the predictive maintenance system.This paper investigates the latest RUL prediction methods,focusing on direct RUL prediction based on machine learning.Firstly,we describe the four representative machine learning models adopted by the RUL prediction methods,including support vector regression(SVR),multilayer perceptron(MLP),convolutional neural network(CNN),and recurrent neural network(RNN).And then,we give the three primary benchmark datasets and two performance evaluation metrics widely used in RUL prediction.The contribution of this paper is to demonstrate the steps and key technical details of how to build the RUL prediction models over the benchmark dataset(C-MAPSS) provided by NASA.We also compare the performance of these representative prediction models in detail and visually analyze the experimental results.Experimental results show that the performance of SVR with a shallow structure is significantly weaker than those based on deep neural networks.CNN and RNN based models have a solid ability for mining complex feature interaction and temporal feature interaction.Finally,we provide an outlook on the future of predictive maintenance technology and discuss the main challenges.

Key words: Predictive maintenance, Remaining useful life prediction, Machine learning, Support vector regression, Multilayer perceptron, Convolutional neural network, Recurrent neural network

中图分类号: 

  • TP393
[1]KOTHAMASU R,HUANG S H,VERDUIN W H.Systemhealth monitoring and prognostics-a review of current paradigms and practices [J].International Journal of Advanced Manufacturing Technology,2006,28(9/10):1012-1024.
[2]ZENG Y Y.Research on life prediction algorithm of Nuclear power plant based on condition monitoring data [D].Beijing:Tsinghua University,2017.
[3]CHEN H Y.Research on condition prediction and diagnosismethod of thermal hydraulic System in nuclear Power plant [D].Heilongjiang:Harbin Engineering University,2018.
[4]REZAEIANJOUYBARI B,YI S.Deep learning for prognostics and health management:State of the art,challenges,and opportunities [J].Measurement,2020,163:107929.
[5]ZHANG L,LIN J,LIU B,et al.A Review on Deep Learning Applications in Prognostics and Health Management [J].IEEE Access,2019,7(1):162415-162438.
[6]SI X S,WANG W,HU C H,et al.Remaining useful life estimation-A review on the statistical data driven approaches [J].European Journal of Operational Research,2011,213(1):1-14.
[7]SAXENA A,KAI G,SIMON D,et al.Damage propagation modeling for aircraft engine run-to-failure simulation[C]//2008 International Conference on Prognostics and Health Management.IEEE,2008.
[8]WANG Y R,YANG S,LI H X,et al.Determination and lifeprediction of fatigue parameters in total strain life equation [J].Chinese Journal of Aerospace Power,2018,33(1):1-14.
[9]WU X R,LIU J Z.Fatigue life prediction of aerospace materials based on small crack theory [J].Chinese Journal of aviation,2006,27(2):219-226.
[10]YANG J H,KOC M,LEE J.A prognostic algorithm for machine performance assessment and its application [J].Production Planning & Control,2007,15(8):796-801.
[11]WANG X L,HAN G,LI X,et al.A SVR-Based Remaining Life Prediction for Rolling Element Bearings [J].Journal of Failure Analysis and Prevention,2015,15(4):548-554.
[12]TIAN Z.An artificial neural network method for remaining useful life prediction of equipment subject to condition monitoring [J].Journal of Intelligent Manufacturing,2012,23(2):227-237.
[13]LI X,DING Q,SUN J Q.Remaining useful life estimation in prognostics using deep convolution neural networks [J].Reliability Engineering & System Safety,2018,172(APR.):1-11.
[14]SONG Y,GAO S,LI Y,et al.Distributed Attention-Based Temporal Convolutional Network for Remaining Useful Life Prediction[J].IEEE Internet of Things Journal,2020,8(12):9594-9602.
[15]SHUAI Z,RISTOVSKI K,FARAHAT A,et al.Long Short-Term Memory Network for Remaining Useful Life estimation[C]//2017 IEEE International Conference on Prognostics and Health Management(ICPHM).IEEE,2017.
[16]LIAO Y,ZHANG L,LIU C.Uncertainty Prediction of Remaining Useful Life Using Long Short-Term Memory Network Based on Bootstrap Method[C]//2018 IEEE International Conference on Prognostics and Health Management (ICPHM).IEEE,2018:1-8.
[17]YU W,KIM I Y,MECHEFSKE C.Remaining useful life estimation using a bidirectional recurrent neural network based autoencoder scheme [J].Mechanical Systems and Signal Proces-sing,2019,129(AUG.15):764-780.
[18]REN L,SUN Y,CUI J,et al.Bearing remaining useful life prediction based on deep autoencoder and deep neural networks [J].Journal of Manufacturing Systems,2018,48:71-77.
[19]LIU X Y,XIONG Z G,YAN C G.A fusion support vector machine-Prediction of residual service life of machinery by high order particle filter method [J].Journal of Guizhou University:Natural Science Edition,2018,35(5):74-80.
[20]HU J,QIAN X,CHENG H,et al.Remaining useful life prediction for aircraft engines based on phase space reconstruction and hybrid VNS-SVR model [J].Journal of Intelligent and Fuzzy Systems,2021,41(2):3415-3428.
[21]RH A,LX A,XL B,et al.Residual life predictions for ball bea-rings based on self-organizing map and back propagation neural network methods [J].Mechanical Systems and Signal Proces-sing,2007,21(1):193-207.
[22]PEEL L.Data driven prognostics using a Kalman filter ensemble of neural network models[C]//2008 International Conference on Prognostics and Health Management.IEEE,2008.
[23]WANG X,WANG T,MING A,et al.Deep Spatiotemporal Convolutional Neural Network-based Remaining Useful Life Estimation of Bearings [J].Chinese Journal of Mechanical Enginee-ring,2021,34:1-15.
[24]BABU G S,ZHAO P,LI X L.Deep Convolutional Neural Network Based Regression Approach for Estimation of Remaining Useful Life [C]//International Conference on Database Systems for Advanced Applications.Cham:Springer,2016.
[25]LU Y W,HSU C Y,HUANG K C.An Autoencoder Gated Recurrent Unit for Remaining Useful Life Prediction[J].Processes,2020,8(9):1-18..
[26]CHEN Z,WU M,ZHAO R,et al.Machine Remaining Useful Life Prediction via an Attention Based Deep Learning Approach [J].IEEE Transactions on Industrial Electronics,2021,68(3):2521-2531.
[27]REN L,CHENG X,WANG X,et al.Multi-scale Dense Gate Recurrent Unit Networks for bearing remaining useful life prediction [J].Future generation computer systems,2019,94(MAY):601-609.
[28]ELSHEIKH A,YACOUT S,OUALI M S.Bidirectional handshaking LSTM for remaining useful life prediction [J].Neurocomputing,2019,323(Jan.5):148-156.
[29]GUGULOTHU N,VISHNU T V,MALHOTRA P,et al.Predicting remaining useful life using time series embeddings based on recurrent neural networks[C]//2nd ML PHM Work.SIGKDD 2017,Halifax,Canada,2017.
[30]CHEN Z Q.Research on equipment health condition evaluation and residual life prediction method based on LSTM network [D].Anhui:University of Science and Technology of China,2019.
[31]AL DULAIMI A,ZABIHI S,ASIF A,et al.A multimodal and hybrid deep neural network model for Remaining Useful Life estimation [J].Computers in Industry,2019,108:186-196.
[32]EKER,Ö F,CAMCI F,JENNIONS I K.Major Challenges in Prognostics:Study on Benchmarking Prognostics Datasets [C]//Proceedings of the European Conference of the PHM Society 2012.PHM Society,2012.
[33]ZHANG L Z.Research on Residual Service Life prediction of Rolling Bearings based on time-frequency analysis and deep learning [D].Shandong:ShandongUniversity,2021.
[34]FAN B.Research on fault prediction method for health manage-ment of aircraft Key Components [D].Hunan:National University of Defense Technology,2015.
[35]LI X,ZHANG W,DING Q.Deep learning-based remaining useful life estimation of bearings using multi-scale feature extraction [J].Reliability Engineering & System Safety,2019,182(FEB.):208-218.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发