基于深度学习模型的滚动轴承剩余使用寿命预测的可靠性评估

Santosh Bisoyi; Amit Kumar Rathi; Swarup Mahato

doi:10.1088/3050-2454/ae61b4

您当前的位置：

首页 >

文章列表页 >

基于深度学习模型的滚动轴承剩余使用寿命预测的可靠性评估

Papers | 更新时间：2026-05-08

- 基于深度学习模型的滚动轴承剩余使用寿命预测的可靠性评估
- Reliability assessment of remaining useful life predictions of roller bearings using deep learning models
- 可靠性科学与工程学报（英文） 2026年页码：1-23
- 作者机构：
  
  Department of Civil and Infrastructure Engineering, Indian Institute of Technology Jodhpur, Jodhpur, Rajasthan 342030, India
  Rishabh Centre Research and Innovation in Clean Energy, Indian Institute of Technology Jodhpur, Jodhpur, Rajasthan 342030, India
  Department of Civil Engineering, Indian Institute of Engineering Science and Technology, Shibpur, West Bengal 711103, India
- 作者简介：
  
  [ "Santosh Bisoyi received a BTech. degree in Civil Engineering from the College of Engineering and Technology, Bhubaneswar, Odisha, India, in 2019 and an MTech degree in Structural Engineering from the National Institute of Technology Agartala, Tripura, India, in 2021. He is currently working toward a PhD degree in Civil and Infrastructure Engineering at the Indian Institute of Technology Jodhpur, Rajasthan, India. His research interests include structural parameter identification, damage assessment, RUL, retrofitting, signal processing, machine learning, and DL." ]
  [ "Amit Kumar Rathi received a BE degree in Civil Engineering from Jai Narain Vyas University, India, in 2009, and an MTech degree from the Indian Institute of Technology Guwahati, India, in 2011. In 2019, he was awarded a PhD in Structural Engineering from the Indian Institute of Technology Guwahati, India. He is presently serving as an Assistant Professor in the Department of Civil and Infrastructure Engineering and Rishabh Centre Research and Innovation in Clean Energy (Secondary Affiliation) at the Indian Institute of Technology Jodhpur, Rajasthan, India. He has held faculty positions at the National Institute of Technology Sikkim, India and National Institute of Technology Calicut, India. His research focuses on reliability analysis and design, uncertainty quantification, stochastic modelling, structural dynamics, health monitoring, condition assessment and resilience." ]
  [ "Swarup Mahato received a BE degree in Civil Engineering from Bengal Engineering & Science University, Shibpur, in 2010, and MTech and PhD degrees in Structural Engineering from the Indian Institute of Technology Guwahati in 2013 and 2019, respectively. He has held postdoctoral and research positions at Université Gustave Eiffel, France; Kaunas University of Technology, Lithuania; and Aarhus University, Denmark, where his work focused on vibration-based monitoring, photogrammetric techniques, and digital twin–based sensing for structural systems. He is currently serving as an Assistant Professor in the Department of Civil Engineering at the Indian Institute of Engineering Science and Technology, Shibpur, India. His research broadly spans structural health monitoring, signal processing for condition assessment and predictive maintenance, time–frequency analysis, wireless sensing networks, vibration control, and bridge dynamics." ]
- 基金信息：
- DOI：10.1088/3050-2454/ae61b4
  中图分类号：
- 收稿：2025-12-31，
  
  修回：2026-03-23，
  
  录用：2026-04-19，
  
  网络首发：2026-04-29，
  
  纸质出版：2026-06
- 稿件说明：
移动端阅览
Santosh Bisoyi, Amit Kumar Rathi and Swarup Mahato. 基于深度学习模型的滚动轴承剩余使用寿命预测的可靠性评估[J]. 可靠性科学与工程学报（英文）, 2026, 2: 025302.

Santosh Bisoyi, Amit Kumar Rathi and Swarup Mahato. Reliability assessment of remaining useful life predictions of roller bearings using deep learning models[J]. Journal of Reliability Science and Engineering, 2026, 2: 025302.
Santosh Bisoyi, Amit Kumar Rathi and Swarup Mahato. 基于深度学习模型的滚动轴承剩余使用寿命预测的可靠性评估[J]. 可靠性科学与工程学报（英文）, 2026, 2: 025302. DOI： 10.1088/3050-2454/ae61b4.

Santosh Bisoyi, Amit Kumar Rathi and Swarup Mahato. Reliability assessment of remaining useful life predictions of roller bearings using deep learning models[J]. Journal of Reliability Science and Engineering, 2026, 2: 025302. DOI： 10.1088/3050-2454/ae61b4.

摘要

本研究聚焦于轴承剩余使用寿命（RUL）的精确可靠预测。近年来，基于机器学习的RUL预测在确保工业机械的安全性、效率和成本效益方面受到关注。然而，从全寿命运行实验中采集的原始振动信号通常具有噪声大、非平稳、高维度等特点，使直接学习退化模式具有挑战性。这常常导致在给定时刻的RUL预测不一致，从而削弱了预测的可靠性。为此，本研究提出了一种可靠的多阶段预测框架，集成了稳健的信号预处理、数据驱动的健康状态分类以及基于深度学习的RUL估计。该流程采用分段聚合近似进行降维，利用奇异谱分析抑制噪声，并引入基于中心矩的变点检测方法自动识别退化的起始点。随后，构建了用于表征退化过程和有效使用寿命的健康指标。采用极端梯度提升分类器，依据振动响应区分轴承的健康运行状态与退化运行状态。之后，提出了一种混合卷积门控循环(CGR)神经网络来预测RUL。该方法在不同工况的多个全寿命运行轴承数据集上进行了验证。相应结果表明，分类器的平均F1分数达到约0.993，且可靠性指标大于2.0，表明其能够清晰识别健康状态。所提出的CGR网络表现出优异的性能，在RUL预测中具有较高的回归精度。将该网络的效能与当前最先进的方法（包括双向长短期记忆循环神经网络、门控循环单元、基于因果扩张卷积的残差密集网络结合通道注意力机制，以及采用Transformer的时间卷积网络）进行了对比评估。总体而言，结果证实所提出的框架能够实现精确、计算高效且考虑可靠性的RUL预测，使其适用于旋转机械的实时状态监测与预测性维护应用。

Abstract

This study focuses on an accurate and reliable prediction of the remaining useful life (RUL) of the bearings. In recent times

machine learning based RUL predictions have gained traction to ensure safety

efficiency

and cost-effectiveness of industrial machinery. However

raw vibration signals acquired from run-to-failure experiments are typically noisy

non-stationary

and high-dimensional

making direct learning of degradation patterns challenging. This often undermines the reliability of the prediction due to inconsistency in the RUL predictions at a given time. This study proposes a reliable multi-stage prognostic framework that integrates robust signal pre-processing

data-driven health-state classification

and deep learning-based RUL estimation. The proposed pipeline employs a piecewise aggregate approximation for dimensionality reduction

a singular spectrum analysis for noise suppression

and a central-moment-based change point detection to automatically identify the onset of degradation. Subsequently

a health index is constructed to characterize the degradation process and useful service life. An extreme gradient boosting classifier is adopted to distinguish between healthy and degraded operating states of the bearing using vibration response. Later

a hybrid convolution with gated recurrent (CGR) neural network is proposed to predict the RUL. The method is validated using multiple run-to-failure bearing datasets with different operating conditions. The corresponding results show that the classifier achieves average

1 score of approx. 0.993 with reliability indices

2.0

indicating a clear identification of health-state. The proposed CG

R network exhibits attractive performance

with high regression accuracy in RUL prediction. The efficacy of the proposed network is also evaluated against state-of-the-art methods

including bidirectional long-short-term memory recurrent neural networks

gated recurrent units

causal dilated convolution-based residual densenet with channel attention and temporal convolutional network with transformer. Overall

the results confirm that the proposed framework enables accurate

computationally efficient

and reliability-informed RUL predictions

making it suitable for real-time condition monitoring and predictive maintenance applications in rotating machinery.

关键词

Keywords

references

Jardine A K S , Lin D and Banjevic D 2006 A review on machinery diagnostics and prognostics implementing condition-based maintenance Mech. Syst. Signal Process. 20 1483 – 510

Huang R , Xi L , Li X , Liu C R , Qiu H and Lee J 2007 Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods Mech. Syst. Signal Process. 21 193 – 207

Zio E 2022 Prognostics and Health Management (PHM): where are we and where do we (need to) go in theory and practice Reliab. Eng. Syst. Saf. 218 108119

Saidi L , Ben Ali J , Bechhoefer E and Benbouzid M 2017 Wind turbine high-speed shaft bearings health prognosis through a spectral Kurtosis-derived indices and SVR Appl. Acoust. 120 1 – 8

Lei Y , He Z and Zi Y 2011 EEMD method and WNN for fault diagnosis of locomotive roller bearings Expert Syst. Appl. 38 7334 – 41

Hu L , Wang L , Chen Y , Hu N and Jiang Y 2022 Bearing fault diagnosis using piecewise aggregate approximation and complete ensemble empirical mode decomposition with adaptive noise Sensors 22 6599

Liu H , Zhou J , Zheng Y , Jiang W and Zhang Y 2018 Fault diagnosis of rolling bearings with recurrent neural network-based autoencoders ISA Trans. 77 167 – 78

Tobon-Mejia D A , Medjaher K , Zerhouni N and Tripot G 2012 A data-driven failure prognostics method based on mixture of gaussians hidden markov models IEEE Trans. Reliab. 61 491 – 503

Yu J 2012 Health condition monitoring of machines based on hidden Markov model and contribution analysis IEEE Trans. Instrum. Meas. 61 2200 – 11

McDonald G L and Zhao Q 2017 Multipoint optimal minimum entropy deconvolution and convolution fix: application to vibration fault detection Mech. Syst. Signal Process. 82 461 – 77

Liu R , Yang B , Zio E and Chen X 2018 Artificial intelligence for fault diagnosis of rotating machinery: a review Mech. Syst. Signal Process. 108 33 – 47

Kankar P K , Sharma S C and Harsha S P 2011 Rolling element bearing fault diagnosis using wavelet transform Neurocomputing 74 1638 – 45

Soualhi A , Medjaher K and Zerhouni N 2015 Bearing health monitoring based on Hilbert–Huang transform, support vector machine and regression IEEE Trans. Instrum. Meas. 64 52 – 62

Eren L , Ince T and Kiranyaz S 2019 A generic intelligent bearing fault diagnosis system using compact adaptive 1D CNN classifier J. Signal Process. Syst. 91 179 – 89

Star M and McKee K 2021 Remaining useful life estimation using neural ordinary differential equations Int. J. Progn. Health Manage. 12 1

Li J , Wang Z , Liu X and Feng Z 2023 Remaining useful life prediction of rolling bearings using GRU-DeepAR with adaptive failure threshold Sensors 23 1144

Nemani V P , Lu H , Thelen A , Hu C and Zimmerman A T 2022 Ensembles of probabilistic LSTM predictors and correctors for bearing prognostics using industrial standards Neurocomputing 491 575 – 96

Rathore M S and Harsha S P 2022 Prognostics analysis of rolling bearing based on bi-directional LSTM and attention mechanism J. Fail. Anal. Prev. 22 704 – 23

Sun B , Zhang J , Hu W , Wei Y , Wu X and Zhang L 2026 Prediction of remaining useful life for rolling bearings based on DCNN-BiLSTM-CBAM with multi-scale feature extraction Eng. Res. Express 8 025513

Guo J , Zhou J , Meng Q , Du B and Zhou S 2026 Federated transfer learning-based model for remaining useful life prediction of rolling bearing across edge clients Meas. Sci. Technol. 37 026007

Bisoyi S , Rathi A K and Mahato S 2025 Fault diagnosis of rolling bearing failures using a multi-stage e-CNN-GRU-SAM network Sci. Rep. 15 33102

Malhotra P , Ramakrishnan A , Anand G , Vig L , Agarwal P and Shroff G 2016 LSTM-based encoder-decoder for multi-sensor anomaly detection (arXiv: 1607.00148)

Zhang D , Zhou S , Zheng Y and Xu X 2025 Review on application of machine vision-based intelligent algorithms in gear defect detection Processes 13 3370

Garcia J , Rios-Colque L , Peña A and Rojas L 2025 Condition monitoring and predictive maintenance in industrial equipment: an NLP-assisted review of signal processing, hybrid models and implementation challenges Appl. Sci. 15 5465

Chang X , Zhang Y , Liu M , Wang T and Wang H 2025 Proactive health management strategies for older adults with motoric cognitive risk syndrome: a scoping review BMC Geriatr. 26 142

Prasad A , Dantreliya C , Chande M , Chauhan V and Rai A 2023 An intelligent fault diagnosis framework based on piecewise aggregate approximation, statistical moments and sparse autoencoder Proc. Inst. Mech. Eng. Pt. O J. Risk Reliab. 237 686 – 702

Gupta M , Wadhvani R and Rasool A 2023 A real-time adaptive model for bearing fault classification and remaining useful life estimation using deep neural network Knowl.-Based Syst. 259 110070

Li J , Ding W , Mao W , Zhang J , Meng Z and Tong K 2024 Causal dilated convolution-based residual DenseNet with channel attention for RUL prediction of rolling bearings Measurement 235 115012

Cao W , Meng Z , Li J , Wu J and Fan F 2025 A remaining useful life prediction method for rolling bearing based on TCN-transformer IEEE Trans. Instrum. Meas. 74 3501309

Sankararaman S 2015 Significance, interpretation and quantification of uncertainty in prognostics and remaining useful life prediction Mech. Syst. Signal Process. 52–53 228 – 47

Zhang L , Lin J , Liu B , Zhang Z , Yan X and Wei M 2019 A review on deep learning applications in prognostics and health management IEEE Access 7 162415 – 38

Baraldi P , Cadini F , Mangili F and Zio E 2013 Model-based and data-driven prognostics under different available information Probabilist. Eng. Mech. 32 66 – 79

Lin Y H and Li G H 2022 A bayesian deep learning framework for RUL prediction incorporating uncertainty quantification and calibration IEEE Trans. Ind. Informat. 18 7274 – 84

Hu X , Tan L and Tang T 2024 M2BIST-SPNet: RUL prediction for railway signaling electromechanical devices J. Supercomput. 80 16744 – 74

Xuan Q L , Adhisantoso Y G , Munderloh M and Ostermann J 2023 Uncertainty-aware remaining useful life prediction for predictive maintenance using deep learning Proc. CIRP 118 116 – 21

Xu W and Zio E 2024 Uncertainty-aware prediction of remaining useful life in complex systems Proc. Annual Conf. PHM Society 2024 vol 16 pp 1 – 4

Zhang D , Dong J , Wang W , Bérenguer C and Zhao Z 2026 A multi-model ensemble and its portability for the prognostics of direct methanol fuel cells under different dynamic operating conditions Reliab. Eng. Syst. Saf. 270 112179

Yang C L , Meles T Y , Yilma A A and Teshome M M 2026 Uncertainty aware predictive maintenance using a hybrid transformer with Monte Carlo Dropout and conformal prediction Ain Shams Eng. J. 17 103992

Zhou J , Zhang H , Tang B , Zheng L and Wang Y 2026 Dynamic reliability informed adaptive task scheduling for multirobot manufacturing system Adv. Eng. Inform. 71 104335

Lei B , Wang Y , Li M , Liu S and Liu W 2026 Remaining useful life prediction with uncertainty quantification based on multi-scale attention-based physics-informed neural network Eng. Res. Express 8 015521

Song X , Liu J and Xia M 2023 Advanced vibration-based fault diagnosis and vibration control methods Sensors 23 7704

Zhang D , Stewart E , Ye J , Entezami M and Roberts C 2020 Roller bearing degradation assessment based on a deep MLP convolution neural network considering outlier regions IEEE Trans. Instrum. Meas. 69 2996 – 3004

Pan Y , Hong R , Chen J , Qin Z and Feng Y 2019 Incipient fault detection of wind turbine large-size slewing bearing based on circular domain Measurement 137 130 – 42

Rocco S and Claudio M 2013 Singular spectrum analysis and forecasting of failure time series Reliab. Eng. Syst. Saf. 114 126 – 36

Liu K , Law S S , Xia Y and Zhu X Q 2014 Singular spectrum analysis for enhancing the sensitivity in structural damage detection J. Sound Vib. 333 392 – 417

Khan M Z and Li D 2024 Dynamic logistic ensembles with recursive probability and automatic subset splitting for enhanced binary classification 2024 IEEE 15th Annual Ubiquitous Computing, Electronics and Mobile Communication Conf. (UEMCON) pp 440 – 7

García-León R A , Navarro-Barrera C J and Afanador-García N 2025 Colombian regulations in the seismic design of reinforced concrete buildings with portal frames: a comparative and bibliometric analysis Buildings 15 4303

Zhang J , Xiao T , Ji J , Zeng P and Cao Z J 2023 Geotechnical Reliability Analysis: Theories, Methods and Algorithms ( Springer Nature )

Matsumoto E Y and Del-Moral-Hernandez E 2016 Improving regression predictions using individual point reliability estimates based on critical error scenarios Inf. Sci. 374 65 – 84

Zhuang L , Wu X , Chow A H F , Ma W , Lam W H K and Wong S C 2025 Reliability-based journey time prediction via two-stream deep learning with multi-source data J. Intell. Transp. Syst. 29 134 – 52

Chang X , Guo J , Qin H , Huang J , Wang X and Ren P 2024 Single-objective and multi-objective flood interval forecasting considering interval fitting coefficients Water Resour. Manage. 38 3953 – 72

Taylor K E 2001 Summarizing multiple aspects of model performance in a single diagram J. Geophys. Res. Atmos. 106 7183 – 92

Wang B , Lei Y , Li N and Li N 2018 A hybrid prognostics approach for estimating remaining useful life of rolling element bearings IEEE Trans. Reliab. 69 401 – 12

Nectoux P , Gouriveau R , Medjaher K , Ramasso E , Chebel-Morello B , Zerhouni N and Varnier C 2012 PRONOSTIA: an experimental platform for bearings accelerated degradation tests Proc. IEEE Int. Conf. Prognostics and Health Management (PHM 2012)

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据