Residual Life Prediction of Rolling Bearings Driven by Digital Twins
Abstract
:1. Introduction
1.1. Physics-Based Model Methods
1.2. Data-Driven Methods
1.3. Hybrid Model-Based Methods
2. Construction of the Rolling Bearing Remaining Useful Life Prediction System
2.1. Establishment of the Rolling Bearing Digital Twin System
2.2. Integration of Digital Twin Model and Prediction Algorithm
3. Remaining Useful Life Prediction of Rolling Bearings
3.1. Modeling of the Rolling Bearing Digital Twin System
3.2. Digital Twin Parameter Update and Degradation Signal Generation
3.3. Establishment of the Rolling Bearing Remaining Useful Life Prediction Model
3.3.1. Determination of the First Fault Point
3.3.2. BO-BI-GRU Modeling Steps
3.4. Ablation Study on Rolling Bearing Remaining Useful Life Prediction
3.4.1. Dataset Selection
3.4.2. First Fault Point Identification
3.4.3. Data Processing and Feature Extraction
3.4.4. Analysis of Prediction Results
4. Digital Twin Signals for Rolling Bearing RUL Prediction Experiments
4.1. Experimental Data and Experimental Setup
4.2. Experimental Results and Analysis
5. Discussion
6. Conclusions
- (1)
- Development of a digital twin-based RUL prediction system. The full-lifecycle digital twin-integrated model uses Modelica as the digital twin entity. Acceleration vibration signals of rolling bearings are selected as the data source for RUL prediction. A hybrid approach is proposed to integrate digital twin technology and data-driven RUL prediction methods.
- (2)
- Establishment of an improved recurrent neural network-based RUL prediction model. An enhanced gated recurrent unit (GRU) model is utilized, with Bayesian optimization applied to fine-tune network hyperparameters. The BO-BI-GRU network is developed as the application-layer prediction model. The model is validated using the XJTU-SY rolling bearing accelerated life dataset from Xi’an Jiaotong University, and both comparative and ablation experiments confirm its effectiveness. The results indicate that the proposed model effectively identifies degradation states in sequential features, enabling accurate RUL prediction for rolling bearings.
- (3)
- Verification of digital twin efficacy in RUL estimation. Comprehensive testing on Paderborn University and XJTU-SY bearing datasets demonstrates the system’s functionality and twin data applicability. The experimental results show that the twin signals exhibit high similarity to actual signals regarding time-domain waveforms and fault characteristic frequencies. A well-balanced combination of actual measured and digital twin data in RUL prediction can optimize data acquisition costs while maintaining model prediction accuracy.
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Lung, B.; Monnin, M.; Voisin, A. Degradation state model-based prognosis for proactively maintaining product performance. CIRP Ann.-Manuf. Technol. 2008, 57, 49–52. [Google Scholar] [CrossRef]
- Fan, B.; Hu, L.; Hu, N.Q. Remaining useful life prediction of rolling bearings by the particle filter method based on degradation rate tracking. J. Vibroengineering 2015, 17, 743–756. [Google Scholar]
- Qian, Y.N.; Yan, R.Q.; Gao, R.X. A multi-time scale approach to remaining useful life prediction in rolling bearing. Mech. Syst. Signal Process. 2017, 83, 549–567. [Google Scholar] [CrossRef]
- Peng, Y.F.; Cheng, J.S.; Liu, Y.F. An adaptive data-driven method for accurate prediction of remaining useful life of rolling bearings. Front. Mech. Eng. 2018, 13, 301–310. [Google Scholar] [CrossRef]
- Gao, X.D.; Shao, Y.B.; Xie, L.Y. Prediction of fatigue crack propagation of X56 steel submarine pipelines in corrosive environment. Mater. Rep. 2020, 34, 2123–2130. [Google Scholar]
- Qu, X.G.; Zhang, X.K.; Jiao, S.Y. Numerical simulation research on residual life of bridge crane based on fracture mechanics. J. Saf. Environ. 2022, 22, 1284–1290. [Google Scholar] [CrossRef]
- Chen, M.Z.; Zhou, D.H.; Liu, G.P. A new particle predictor for fault prediction of nonlinear time-varying systems. Dev. Chem. Eng. Miner. Process. 2010, 13, 379–388. [Google Scholar] [CrossRef]
- Tobon-Mejia, D.A.; Medjaher, K.; Zerhouni, N. A data-driven failure prognostics method based on mixture of Gaussians hidden Markov models. IEEE Trans. Reliab. 2012, 61, 491–503. [Google Scholar] [CrossRef]
- Wu, Q.; Zhang, C.S. Cascade fusion convolutional long-short time memory network for remaining useful life prediction of rolling bearing. IEEE Access 2020, 8, 32957–32965. [Google Scholar] [CrossRef]
- Pan, Z.Z.; Meng, Z.; Chen, Z.J. A two-stage method based on extreme learning machine for predicting the remaining useful life of rolling-element bearings. Mech. Syst. Signal Process. 2020, 144, 106899. [Google Scholar] [CrossRef]
- Zhao, H.M.; Liu, H.D.; Jin, Y. Feature extraction for data-driven remaining useful life prediction of rolling bearings. IEEE Trans. Instrum. Meas. 2021, 70, 3511910. [Google Scholar] [CrossRef]
- Sun, S.Y.; Zhang, G.; Liang, W.G.; She, B.; Tian, F.Q. Remaining useful life prediction method of rolling bearing based on time series data augmentation and BLSTM. Syst. Eng. Electron. 2022, 44, 1060–1068. [Google Scholar] [CrossRef]
- Wang, F.F.; Tang, S.J.; Sun, X.Y.; Qi, S.; Yu, C.Q.; Si, X.S. Remaining useful life prediction based on multi-source information with considering random effects. J. Beijing Univ. Aeronaut. Astronaut. 2023, 49, 3075–3085. [Google Scholar] [CrossRef]
- Cao, Y.D.; Ding, Y.F.; Jia, M.P. A novel temporal convolutional network with residual self-attention mechanism for remaining useful life prediction of rolling bearings. Reliab. Eng. Syst. Saf. 2021, 215, 107813. [Google Scholar] [CrossRef]
- Zhuang, J.C.; Jia, M.P.; Zhao, X.L. An adversarial transfer network with supervised metric for remaining useful life prediction of rolling bearing under multiple working conditions. Reliab. Eng. Syst. Saf. 2022, 225, 108599. [Google Scholar] [CrossRef]
- Eknat, K.G.; Diwakar, G. Prediction of remaining useful life of rolling bearing using hybrid DCNN-BiGRU model. J. Vib. Eng. Technol. 2023, 11, 997–1010. [Google Scholar] [CrossRef]
- Guo, J.Y.; Wang, J.; Wang, Z.Y. A CNN-BiLSTM-Bootstrap integrated method for remaining useful life prediction of rolling bearings. Qual. Reliab. Eng. Int. 2023, 39, 1796–1813. [Google Scholar] [CrossRef]
- Han, Y.; Song, X.P.; Shi, J.M. KPCA-WPHM-SCNs-based remaining useful life prediction method for motor rolling bearings. Trans. Inst. Meas. Control 2024, 46, 973–991. [Google Scholar] [CrossRef]
- Cao, H.J.; Xiao, W.; Sun, J. A hybrid data- and model-driven learning framework for remaining useful life prognostics. Eng. Appl. Artif. Intell. 2024, 135, 108557. [Google Scholar] [CrossRef]
- Zhang, Y.Z.; Zhao, X.Q.; Xu, R.R. A dual-stream temporal convolutional network for remaining useful life prediction of rolling bearings. Meas. Sci. Technol. 2025, 36, 016206. [Google Scholar] [CrossRef]
- Qin, Y.; Liu, H.; Mao, Y. Faulty rolling bearing digital twin model and its application in fault diagnosis with imbalanced samples. Adv. Eng. Inform. 2024, 61, 102513. [Google Scholar] [CrossRef]
- Cui, L.; Xiao, Y.; Liu, D.; Han, H. Digital twin-driven graph domain adaptation neural network for remaining useful life prediction of rolling bearing. Reliab. Eng. Syst. Saf. 2024, 245, 109991. [Google Scholar] [CrossRef]
- Lei, Y.G.; Han, T.Y.; Wang, B. XJTU-SY rolling element bearing accelerated life test datasets: A tutorial. J. Mech. Eng. 2019, 55, 1–6. [Google Scholar] [CrossRef]
- Li, N.P.; Wang, M.Y.; Si, X.S. Remaining useful life prediction of lithium-ion battery with nonparametric degradation modeling and incomplete data. Reliab. Eng. Syst. Saf. 2025, 256, 110721. [Google Scholar] [CrossRef]
- Wang, Y.; Lei, Y.G.; Jiang, J.Z. A multimodal dynamic parameterized bilinear factorized framework for remaining useful life prediction under variational data. Reliab. Eng. Syst. Saf. 2024, 245, 110025. [Google Scholar] [CrossRef]
- Bharatheedasan, K.; Maity, T.; Durairaj, M. Enhanced fault diagnosis and remaining useful life prediction of rolling bearings using a hybrid multilayer perceptron and LSTM network model. Alex. Eng. J. 2025, 115, 355–369. [Google Scholar] [CrossRef]
- Zhang, Z.H.; Zhang, W.; Cai, J. Remaining life prediction of rolling bearings based on unsupervised anomaly detection and STA-BiLSTM. IEEE Sens. J. 2024, 24, 41659–41674. [Google Scholar] [CrossRef]
Symbol | Meaning | Value | Symbol | Meaning | Value |
---|---|---|---|---|---|
D0 | Bearing outside diameter | 39.8 mm | Di | Bearing bore diameter | 29.3 mm |
D | Bearing pitch diameter | 34.55 mm | Db | Rolling diameter | 7.92 mm |
B | Bearing width | 15 mm | nb | Number of rolling elements | 8 |
α | Contact angle | 0° | C | Basic dynamic load rating | 12,820 N |
C0 | Basic static load rating | 6650 N |
Feature Index | Feature Calculation Method | Feature Index | Feature Calculation Method |
---|---|---|---|
Mean value | Peak value | ||
Maximum value | Skewness | ||
Minimum value | Waveform index | ||
Variance | Peak index | ||
Root amplitude | Pulse index | ||
Absolute average amplitude | Kurtosis index | ||
RMS value | Margin index | ||
Kurtosis | Skewness |
Model | RMSE | MAE | R2 |
---|---|---|---|
BP | 0.6374 | 0.6691 | 0.4576 |
SVR | 0.4551 | 0.4624 | 0.5214 |
CNN-BI-GRU | 0.2414 | 0.2725 | 0.6872 |
BO-BI-GRU | 0.2147 | 0.2362 | 0.7191 |
Model | RMSE | MAE | R2 |
---|---|---|---|
RNN | 0.3953 | 0.4285 | 0.5501 |
GRU | 0.3112 | 0.3459 | 0.6085 |
BI-GRU | 0.2618 | 0.2872 | 0.6624 |
BO-BI-GRU | 0.2147 | 0.2362 | 0.7191 |
Experimental Bearing | Data Combination | RMSE | MAE | R2 |
---|---|---|---|---|
1_3 | 100% twin data | 0.2735 | 0.3122 | 0.6485 |
50% measured + 50% twin | 0.2504 | 0.2814 | 0.6760 | |
100% measured data | 0.2421 | 0.2739 | 0.7065 | |
3_1 | 100% twin data | 0.2788 | 0.3143 | 0.6392 |
50% measured + 50% twin | 0.2539 | 0.2764 | 0.6694 | |
100% measured data | 0.2473 | 0.2703 | 0.6983 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Fan, J.; Zhao, L.; Li, M. Residual Life Prediction of Rolling Bearings Driven by Digital Twins. Symmetry 2025, 17, 406. https://doi.org/10.3390/sym17030406
Fan J, Zhao L, Li M. Residual Life Prediction of Rolling Bearings Driven by Digital Twins. Symmetry. 2025; 17(3):406. https://doi.org/10.3390/sym17030406
Chicago/Turabian StyleFan, Jiayi, Lijuan Zhao, and Minghao Li. 2025. "Residual Life Prediction of Rolling Bearings Driven by Digital Twins" Symmetry 17, no. 3: 406. https://doi.org/10.3390/sym17030406
APA StyleFan, J., Zhao, L., & Li, M. (2025). Residual Life Prediction of Rolling Bearings Driven by Digital Twins. Symmetry, 17(3), 406. https://doi.org/10.3390/sym17030406