.png){kind=icon}
.png){kind=icon}
Return
A hybrid deep learning-based framework for rolling bearing fault diagnosis: Multi-resolution feature extraction and enhanced adaptive nonlinear mapping
.png){kind=icon}
PRE
AI
W
Z
N
K
DOI：10.1016/j.advengsoft.2026.104184.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Proposes a staged MSCNN–KAN–Transformer–LSTM framework for bearing fault diagnosis. • Introduces KAN-based spline mapping for adaptive nonlinear feature reshaping. • Validates the framework through ablation, sensitivity, repeated-run, and cross-condition tests. • Provides interpretable evidence via class-wise analysis and layer-wise feature evolution. • Achieves strong mean accuracies on the CWRU and SEU bearing datasets.
Keywords:
bearing fault diagnosis
MSCNN–KAN–Transformer–LSTM
adaptive nonlinear mapping
multi-resolution feature extraction
interpretable deep learning
Journal
IF:
5.7
Papers:
3.3K
Citations:
1.2W
Organization
Cited Papers