Fault Diagnosis of Rotating Machinery Based on One-Dimensional Deep Residual Shrinkage Network with a Wide Convolution Layer

Table 6

The network parameters of the 1D-WDRSN model.

No.	Layer type	Kernel size	Kernel number	Output size (Width×Depth)
1	Conv1d_1	256 × 1	48	512 × 48
2	RBU_1_Conv_1	16 × 1	48	256 × 48
3	RBU_1_Conv_2	16 × 1	48	256 × 48
4	RBU_2_Conv_1	16 × 1	48	128 × 48
5	RBU_2_Conv_2	16 × 1	48	128 × 48
6	Softmax	10	1	10