New Method for Diagnosing Rolling Bearing Faults Using AI
Author Information
Author(s): Wang Chaobing, Huang Cong, Zhang Long, Xiang Zhibin, Xiao Yiwen, Qian Tongshuai, Liu Jiayang
Primary Institution: East China Jiaotong University
Hypothesis
Can combining denoising diffusion implicit models with a convolutional neural network improve fault diagnosis accuracy for rolling bearings under imbalanced data conditions?
Conclusion
The proposed method achieves over 99% recognition accuracy on two datasets, significantly outperforming existing fault diagnosis methods.
Supporting Evidence
- The method effectively mitigates the sample imbalance issue and reduces the risk of overfitting during small-sample training.
- The proposed model provides better generalization performance and recognition accuracy than existing fault diagnosis methods.
- Experimental results demonstrate that the proposed method achieves over 99% recognition accuracy on both the CWRU dataset and the Nanchang Railway Bureau dataset.
Takeaway
This study shows a new way to find problems in machine parts by using smart computer models that learn from examples, even when there aren't many examples to learn from.
Methodology
The study uses a combination of denoising diffusion implicit models and a convolutional neural network framework to diagnose faults in rolling bearings.
Limitations
The study primarily focuses on fault diagnosis under a single operating condition.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website