journal article·2025-06-27Open access

Analysis of Gearbox Bearing Fault Diagnosis Method Based on 2D Image Transformation and 2D-RoPE Encoding

Xudong Luo, Minghui Wang, Zhijie Zhang

Applied Sciences· Volume 15· Issue 13

Abstract

The stability of gearbox bearings is crucial to the operational efficiency and safety of industrial equipment, as their faults can lead to downtime, economic losses, and safety risks. Traditional models face difficulties in handling complex industrial time-series data due to insufficient feature extraction capabilities and poor training stability. Although transformers show advantages in fault diagnosis, their ability to model local dependencies is limited. To improve feature extraction from time-series data and enhance model robustness, this paper proposes an innovative method based on the ViT. Time-series data were converted into two-dimensional images using polar coordinate transformation and Gramian matrices to enhance classification stability. A lightweight front-end encoder and depthwise feature extractor, combined with multi-scale depthwise separable convolution modules, were designed to enhance fine-grained features, while two-dimensional rotary position encoding preserved temporal information and captured temporal dependencies. The constructed RoPE-DWTrans model implemented a unified feature extraction process, significantly improving cross-dataset adaptability and model performance. Experimental results demonstrated that the RoPE-DWTrans model achieved excellent classification performance on the combined MCC5 and HUST gearbox datasets. In the fault category diagnosis task, classification accuracy reached 0.953, with precision at 0.959, recall at 0.973, and an F1 score of 0.961; in the fault category and severity diagnosis task, classification accuracy reached 0.923, with precision at 0.932, recall at 0.928, and an F1 score of 0.928. Compared with existing methods, the proposed model showed significant advantages in robustness and generalization ability, validating its effectiveness and application potential in industrial fault diagnosis.
Source: crossref · Quality: baseline

Research context

OpenAlex topics: Machine Fault Diagnosis Techniques · Image and Object Detection Techniques · Advanced Measurement and Detection Methods

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

College of Biomedical Engineering, Sichuan University, Chengdu 610065, China

Institute of Regulatory Science for Medical Devices, Sichuan University, Chengdu 610065, China

Funding

National Key Research and Development Program of China

References (53)

Show source-provided references
1. Reliability analysis of rolling ball bearings considering the bearing radial clearance and operating temperature10.46793/adeletters.2022.1.1.3 ↗
2. Fault analysis of gearboxes in open pit mine10.18485/aeletters.2020.5.2.3 ↗
3. Molęda, M., Małysiak, M., Sunderam, V., Ding, W., and Mrozek, D. (2023). From corrective to predictive maintenance—A review of maintenance approaches for the power industry. Sensors, 23.10.3390/s23135970 ↗
4. Shao, Z., Zhang, T., and Kosasih, B. (2024, January 11–13). Compound Faults Diagnosis in Wind Turbine Gearbox Based on Deep Learning Methods: A Review. Proceedings of the 2024 Global Reliability and Prognostics and Health Management Conference (PHM-Beijing), Beijing, China.10.1109/PHM-Beijing63284.2024.10874503 ↗
5. Seo, M., and Yun, W. (2024). Gearbox Condition Monitoring and Diagnosis of Unlabeled Vibration Signals Using a Supervised Learning Classifier. Machines, 12.10.3390/machines12020127 ↗
6. Operational excellence in total productive maintenance: Statistical reliability as support for planned maintenance pillar10.1108/IJQRM-09-2023-0290 ↗
7. Machine learning-based predictive maintenance for fault detection in rotating machinery: A case study10.48084/etasr.6813 ↗
8. Enhancing manufacturing efficiency and quality through automation and deep learning: Addressing redundancy, defects, vibration analysis, and material strength optimization10.30574/wjarr.2024.23.3.2800 ↗
9. Multi-sensor fusion fault diagnosis method of wind turbine bearing based on adaptive convergent viewable neural networks10.1016/j.ress.2024.109980 ↗
10. A literature review of fault diagnosis based on ensemble learning10.1016/j.engappai.2023.107357 ↗
11. Xu, L., Teoh, S., and Ibrahim, H. (2024). A deep learning approach for electric motor fault diagnosis based on modified InceptionV3. Sci. Rep., 14.10.1038/s41598-024-63086-9 ↗
12. Szegedy, C., Vanhoucke, V., and Ioffe, S. (2016, January 27–30). Rethinking the inception architecture for computer vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.10.1109/CVPR.2016.308 ↗
13. Hu, J., Shen, L., and Sun, G. (2018, January 18–23). Squeeze-and-excitation networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.10.1109/CVPR.2018.00745 ↗
14. Support-vector networks10.1007/BF00994018 ↗
15. Harnessing attention mechanisms in a comprehensive deep learning approach for induction motor fault diagnosis using raw electrical signals10.1016/j.engappai.2023.107643 ↗
16. O’shea, K., and Nash, R. (2015). An introduction to convolutional neural networks. arXiv.
17. Recurrent neural networks
18. A high-performance rolling bearing fault diagnosis method based on adaptive feature mode decomposition and Transformer10.1016/j.apacoust.2024.110156 ↗
19. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., and Polosukhin, I. (2017). Attention is all you need. Adv. Neural Inf. Process. Syst., 30.
20. Singh, A., Mousavi, S., and Gaurav, K. (2024). SHS: Scorpion Hunting Strategy Swarm Algorithm. arXiv.
21. FFT-trans: Enhancing robustness in mechanical fault diagnosis with Fourier transform-based transformer under noisy conditions10.1109/TIM.2024.3381688 ↗
22. Fast Fourier transforms: A tutorial review and a state of the art10.1016/0165-1684(90)90158-U ↗
23. Xie, S., Zhou, S., Sakurada, K., Ishikawa, R., Onishi, M., and Oishi, T. (2024). G2fR: Frequency Regularization in Grid-Based Feature Encoding Neural Radiance Fields. European Conference on Computer Vision, Springer Nature.10.1007/978-3-031-72670-5_11 ↗
24. A sound-vibration physical-information fusion constraint-guided deep learning method for rolling bearing fault diagnosis10.1016/j.ress.2024.110556 ↗
25. Activated gradients for deep neural networks10.1109/TNNLS.2021.3106044 ↗
26. Surprising instabilities in training deep networks and a theoretical analysis
27. Are transformers effective for time series forecasting
28. Rethinking attention mechanism in time series classification10.1016/j.ins.2023.01.093 ↗
29. Temporal signals to images: Monitoring the condition of industrial assets with deep learning image processing algorithms
30. Yi, K., Zhang, Q., and Cao, L. (2023). A survey on deep learning-based time series analysis with frequency transformation. arXiv.
31. Qiu, S., Cui, X., Ping, Z., Shan, N., Li, Z., Bao, X., and Xu, X. (2023). Deep learning techniques in intelligent fault diagnosis and prognosis for industrial systems: A review. Sensors, 23.10.3390/s23031305 ↗
32. Wu, G., Ji, X., Yang, G., Jia, Y., and Cao, C. (2023). Signal-to-image: Rolling bearing fault diagnosis using ResNet family deep-learning models. Processes, 11.10.3390/pr11051527 ↗
33. Li, Z., Fan, R., and Tu, J. (2024). Tdanet: A novel temporal denoise convolutional neural network with attention for fault diagnosis. arXiv.
34. Fault diagnosis of rolling bearing based on empirical mode decomposition and improved manhattan distance in symmetrized dot pattern image10.1016/j.ymssp.2021.107817 ↗
35. Bearing fault diagnosis based on Gramian angular field and DenseNet10.3934/mbe.2022656 ↗
36. Wang, M., Wang, W., Zhang, X., and Iu, H.H.-C. (2022). A new fault diagnosis of rolling bearing based on Markov transition field and CNN. Entropy, 24.10.3390/e24060751 ↗
37. Feature extraction of multi-sensors for early bearing fault diagnosis using deep learning based on minimum unscented kalman filter10.1016/j.engappai.2023.107138 ↗
38. New extension of the Kalman filter to nonlinear systems10.1117/12.280797 ↗
39. An ensemble deep learning network based on 2D convolutional neural network and 1D LSTM with self-attention for bearing fault diagnosis10.1016/j.asoc.2025.112889 ↗
40. Long Short-Term Memory10.1162/neco.1997.9.8.1735 ↗
41. Dosovitskiy, A., Beyer, L., and Kolesnikov, A. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. arXiv.
42. Heo, B., Park, S., Han, D., and Yun, S. (2024). Rotary position embedding for vision transformer. European Conference on Computer Vision, Springer Nature.10.1007/978-3-031-72684-2_17 ↗
43. A graph representation learning-based method for fault diagnosis of rotating machinery under time-varying speed conditions10.1007/s11071-025-11151-x ↗
44. Multi-mode Fault Diagnosis Datasets of Gearbox Under Variable Working Conditions10.1016/j.dib.2024.110453 ↗
45. Domain generalization for cross-domain fault diagnosis: An application-oriented perspective and a benchmark study10.1016/j.ress.2024.109964 ↗
46. Chollet, F. (2017, January 21–26). Xception: Deep learning with depthwise separable convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.10.1109/CVPR.2017.195 ↗
47. Ramachandran, P., Zoph, B., and Le, Q. (2017). Searching for activation functions. arXiv.
48. Roformer: Enhanced transformer with rotary position embedding10.1016/j.neucom.2023.127063 ↗
49. Rezero is all you need: Fast convergence at large depth
50. Multireceptive field graph convolutional networks for machine fault diagnosis
51. Attention-aware temporal–spatial graph neural network with multi-sensor information fusion for fault diagnosis10.1016/j.knosys.2023.110891 ↗
52. Research on gearbox fault diagnosis method under variable working conditions based on HHO-MLP neural network
53. Model-assisted multi-source fusion hypergraph convolutional neural networks for intelligent few-shot fault diagnosis to electro-hydrostatic actuator10.1016/j.inffus.2023.102186 ↗