Research Article
A Deep Domain-Adversarial Transfer Fault Diagnosis Method for Rolling Bearing Based on Ensemble Empirical Mode Decomposition
Table 5
Fault classification results using various network models.
| | Working condition | EMBRN | EMBRNDN | EMBRNMD | EMBRNDNMD |
| | A->B | 87.33% | 94.83% | 94.33% | 97.83% | | A->C | 78.50% | 92.17% | 93.50% | 97.33% | | A->D | 73.17% | 95.17% | 94.33% | 97.50% | | B->A | 85.33% | 94.33% | 94.50% | 96.67% | | B->C | 94.83% | 98.33% | 100% | 100% | | B->D | 82.17% | 97.50% | 95.17% | 98.17% | | C->A | 79.33% | 93.50% | 93.83% | 96.50% | | C->B | 90.50% | 97.33% | 97.17% | 98.33% | | C->D | 94.33% | 98.17% | 98.33% | 99.50% | | D->A | 75.32% | 92.17% | 94.83% | 97.17% | | D->B | 81.50% | 95.83% | 97.17% | 98.33% | | D->C | 93.17% | 97.33% | 98.83% | 99.83% | | Average | 84.50% | 95.67% | 95.83% | 98.17% |
|
|
