One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions
Abstract In industrial settings, bearing health directly affects equipment stability, making accurate and efficient fault diagnosis critical for operational safety. Recently, Transformer models have been widely adopted in bearing fault diagnosis due to their strong global modeling capabilities. Howe...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-12533-2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849235494111543296 |
|---|---|
| author | Shaobin Cai Yuchen Wang Wanchen Cai Yuchang Mo Liansuo Wei |
| author_facet | Shaobin Cai Yuchen Wang Wanchen Cai Yuchang Mo Liansuo Wei |
| author_sort | Shaobin Cai |
| collection | DOAJ |
| description | Abstract In industrial settings, bearing health directly affects equipment stability, making accurate and efficient fault diagnosis critical for operational safety. Recently, Transformer models have been widely adopted in bearing fault diagnosis due to their strong global modeling capabilities. However, they still face significant challenges under strong noise and limited data. To address this, this paper proposes an end-to-end Vision Transformer with time–frequency fusion and dual attention across spatial and channel dimensions. The model adopts a dual-branch design: the time-domain branch incorporates spatial and channel attention to capture both local and global features, while the frequency-domain branch uses FFT to extract spectral information and fuses it with temporal features for efficient multi-scale modeling. To further enhance sensitivity to local patterns and periodic variations, a cross-scale convolution module and a periodic feedforward network are introduced. Experiments on the CWRU and PU datasets demonstrate that the proposed model achieves 99.42% and 98.14% accuracy, respectively, under noisy and data-scarce conditions. The results confirm superior noise robustness and diagnostic performance over recent state-of-the-art methods, highlighting its practical potential for real-world industrial applications. |
| format | Article |
| id | doaj-art-def2d127590d4f5cb88890078f6c8ca3 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-def2d127590d4f5cb88890078f6c8ca32025-08-20T04:02:45ZengNature PortfolioScientific Reports2045-23222025-07-0115112910.1038/s41598-025-12533-2One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditionsShaobin Cai0Yuchen Wang1Wanchen Cai2Yuchang Mo3Liansuo Wei4Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of ChinaCollege of Information Engineering, Huzhou UniversityCollege of Management, National Taiwan UniversityCollege of Mathematics, Huaqiao UniversityCollege of Mathematics, SuQian UniversityAbstract In industrial settings, bearing health directly affects equipment stability, making accurate and efficient fault diagnosis critical for operational safety. Recently, Transformer models have been widely adopted in bearing fault diagnosis due to their strong global modeling capabilities. However, they still face significant challenges under strong noise and limited data. To address this, this paper proposes an end-to-end Vision Transformer with time–frequency fusion and dual attention across spatial and channel dimensions. The model adopts a dual-branch design: the time-domain branch incorporates spatial and channel attention to capture both local and global features, while the frequency-domain branch uses FFT to extract spectral information and fuses it with temporal features for efficient multi-scale modeling. To further enhance sensitivity to local patterns and periodic variations, a cross-scale convolution module and a periodic feedforward network are introduced. Experiments on the CWRU and PU datasets demonstrate that the proposed model achieves 99.42% and 98.14% accuracy, respectively, under noisy and data-scarce conditions. The results confirm superior noise robustness and diagnostic performance over recent state-of-the-art methods, highlighting its practical potential for real-world industrial applications.https://doi.org/10.1038/s41598-025-12533-2TransformerBearing Fault DiagnosisDeep LearningSelf-attentionCross-scale Convolution |
| spellingShingle | Shaobin Cai Yuchen Wang Wanchen Cai Yuchang Mo Liansuo Wei One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions Scientific Reports Transformer Bearing Fault Diagnosis Deep Learning Self-attention Cross-scale Convolution |
| title | One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| title_full | One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| title_fullStr | One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| title_full_unstemmed | One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| title_short | One-dimensional time-frequency dual-channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| title_sort | one dimensional time frequency dual channel visual transformer for bearing fault diagnosis under strong noise and limited data conditions |
| topic | Transformer Bearing Fault Diagnosis Deep Learning Self-attention Cross-scale Convolution |
| url | https://doi.org/10.1038/s41598-025-12533-2 |
| work_keys_str_mv | AT shaobincai onedimensionaltimefrequencydualchannelvisualtransformerforbearingfaultdiagnosisunderstrongnoiseandlimiteddataconditions AT yuchenwang onedimensionaltimefrequencydualchannelvisualtransformerforbearingfaultdiagnosisunderstrongnoiseandlimiteddataconditions AT wanchencai onedimensionaltimefrequencydualchannelvisualtransformerforbearingfaultdiagnosisunderstrongnoiseandlimiteddataconditions AT yuchangmo onedimensionaltimefrequencydualchannelvisualtransformerforbearingfaultdiagnosisunderstrongnoiseandlimiteddataconditions AT liansuowei onedimensionaltimefrequencydualchannelvisualtransformerforbearingfaultdiagnosisunderstrongnoiseandlimiteddataconditions |