Research Article
An Intelligent Fault Diagnosis Based on Adversarial Generating Module and Semi-supervised Convolutional Neural Network
Table 2
The distribution of dataset.
| No. | Fault pattern | Total size | Training set | Testing set | Labeled samples | Unlabeled samples |
| C1 | Health | 5001024 | 400 | 100 | 25 | 375 | C2 | Gear burr | 5001024 | 400 | 100 | 25 | 375 | C3 | Gear error | 5001024 | 400 | 100 | 25 | 375 | C4 | Gear hard point | 5001024 | 400 | 100 | 25 | 375 | C5 | Gear tooth broken | 5001024 | 400 | 100 | 25 | 375 | C6 | Gear interference | 5001024 | 400 | 100 | 25 | 375 | C7 | Misalignment | 5001024 | 400 | 100 | 25 | 375 | C8 | Gear crack | 5001024 | 400 | 100 | 25 | 375 | Total | ā | 40001024 | 3200 | 800 | 200 | 3000 |
|
|