Fault Diagnosis for High-Speed Train Axle-Box Bearing Using Simplified Shallow Information Fusion Convolutional...

Area of Science:

• Mechanical Engineering
• Artificial Intelligence
• Signal Processing

Background:

• Axle-box bearings are critical components in high-speed trains, requiring reliable fault diagnosis.
• Vibration signals from these bearings are complex, exhibiting nonlinear and nonstationary characteristics.
• Existing deep learning methods for bearing fault diagnosis need enhanced accuracy and efficiency for real-time applications.

Purpose of the Study:

• To propose a novel deep learning approach for accurate and rapid axle-box bearing fault identification.
• To enhance the efficiency and accuracy of bearing fault diagnosis in high-speed trains.

Main Methods:

• A simplified shallow information fusion-convolutional neural network (SSIF-CNN) model was developed.
• Time and frequency domain features were extracted from vibration signals.
• Global convolution transformed feature maps into sequences, which were concatenated for fault identification.

Main Results:

• The SSIF-CNN model demonstrated effective compression of training time.
• The proposed method significantly improved fault diagnosis accuracy compared to a general CNN.
• Experimental results validated the effectiveness of the SSIF-CNN for bearing fault identification.

Conclusions:

• The SSIF-CNN offers a promising solution for high-speed train axle-box bearing fault diagnosis.
• This approach enhances diagnostic accuracy and operational efficiency, meeting real-time requirements.
• The method provides a reliable tool for maintaining the integrity of critical train components.