Research Article
Fault Detection of Reciprocating Compressor Valve Based on One-Dimensional Convolutional Neural Network
Table 3
Parameters of the 1DCNN model.
| Layer (type) | Number of convolution kernels | Output shape | Parameter |
| Input layer | | 121 | 0 | Conv1 (Conv1D) | 32 | 10132 | 128 | Activation ReLU | | 10132 | 0 | MaxPool1 | | 5132 | 0 | Conv2 (Conv1D) | 64 | 3164 | 6208 | Activation ReLU | | 3164 | 0 | MaxPool2 | | 2164 | 0 | FC1/flatten | | 128 | 0 | FC2 | | 64 | 8256 | Softmax | | 7 | 455 |
|
|