一种基于KA-Transformer的轴承声振融合跨工况故障诊断方法
By fusing vibration and acoustic signals using KA-Transformer and employing grouped rational KA encoders and cross-domain alignment strategies, the problems of insufficient fault feature extraction and cross-condition diagnostic robustness of bearings under complex operating conditions are solved, achieving higher fault identification accuracy and stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV OF SCI & TECH
- Filing Date
- 2026-05-08
- Publication Date
- 2026-07-17
AI Technical Summary
Existing bearing fault diagnosis methods are insufficient in their ability to extract fault features under complex working conditions. The feature distribution varies greatly across working conditions, and fixed activation functions are difficult to adaptively represent non-stationary features, resulting in insufficient diagnostic information and insufficient robustness.
A cross-condition fault diagnosis method for bearings based on KA-Transformer is adopted. Vibration and acoustic signals are collected, frequency domain transformed, divided into sub-blocks and embedded. Feature extraction is performed using a weight-shared group rational KA fusion encoder. Combined with gradient inversion mechanism and cross-domain alignment strategy, joint training is carried out to achieve cross-condition fault diagnosis.
It improves the ability to identify weak fault features, enhances the ability to represent non-stationary features and fine-grained fault features, reduces the differences in feature distribution between different operating conditions, and improves the accuracy and robustness of fault diagnosis.
Smart Images
Figure CN122132820B_ABST