Research Article
Bearing Remaining Useful Life Prediction Based on AdCNN and CWGAN under Few Samples
Table 2
Details of the predictor networks in this experiment.
| | Layer type | Kernel | Channel | Stride | Activation | Number of training parameters |
| | Input | — | 1 | — | — | — | | Conv_1 | 21 1 | 32 | 2 | ReLu | 704 | | Pool_1 | 2 1 | 32 | 1 | — | — | | Conv_2 | 11 × 1 | 48 | 2 | ReLu | 576 | | Pool_2 | 2 1 | 48 | 1 | — | — | | Conv_3 | 7 × 1 | 64 | 2 | ReLu | 512 | | Pool_3 | 2 1 | 64 | 1 | — | — | | FC_1 | — | 1 | — | ReLu | 4981248 | | FC_2 | — | 1 | — | ReLu | 65664 | | Output | — | 1 | — | Sigmoid | 257 |
|
|
