Fault diagnosis of CNC machine tool spindle based on SSA-CNN-Transformer
Aiming at the fault diagnosis of CNC machine tool spindle bearing, a fault diagnosis method of CNC machine tool spindle bearing based on SSA-CNN-Transformer is proposed. First, the vibration signal is converted into time-frequency image through continuous wavelet transform, and then the local features of the time-frequency graph of the vibration signal are extracted by using CNN. The attention mechanism of Transformer is used to capture the global dependence of the time-frequency graph of the vibration data. The SSA algorithm is used to optimize the hyperparameters of the model, which improves the generalization ability and precision of fault diagnosis. The results show that this method performs well in the fault diagnosis of spindle bearing of CNC machine tool, with an average diagnosis accuracy of 94.6%.
Authors
- Yang Wang (ORCID: https://orcid.org/0000-0002-4984-0831)
- Baoyan Zhao
- Zhichao Sun
- Yiwen Zhang (ORCID: https://orcid.org/0000-0002-7304-9820)
- Xiaobo Liu (ORCID: https://orcid.org/0000-0001-5921-8491)
- Laurent Li (ORCID: https://orcid.org/0000-0002-3855-3976)
- Xinming Zhang (ORCID: https://orcid.org/0000-0002-6713-1430)
- Chunfang Li (ORCID: https://orcid.org/0000-0002-4673-2800)
- Linsen Song (ORCID: https://orcid.org/0000-0002-2933-2549)
- Anqi Xiong
Institutions
- Changchun University of Science and Technology (CN)
- Foshan University (CN)
- Great Wall Motors (China) (CN)
- First Automotive Works (China) (CN)
- Jilin Province Science and Technology Department (CN)
- Jilin Business and Technology College
- Changchun University (CN)
Publication Details
- Published
- 2024-12-06
- DOI
- https://doi.org/10.1109/ricai64321.2024.10910983
- Citations
- 1
- Primary Topic
- Advanced machining processes and optimization
- Type
- article
- Field-Weighted Citation Impact
- 0.21
Funders
- Natural Science Foundation of Chongqing
- Education Department of Jilin Province