Research Article
Deep Learning Enabled Fault Diagnosis Using Time-Frequency Image Analysis of Rolling Element Bearings
Table 28
Prediction accuracies for CWR.
| | Model | 20 epochs with
early stopping | 30 Epochs with no early stopping | No overlap |
| | Architecture 2 | 75.2% | 86.7% | 67.2% | | Architecture 1 | 90.4% | 95.7% | 87.2% | | Proposed CNN architecture | 83.1% | 98.5% | 93.6% |
|
|
