Bearing Fault Diagnosis Based on Frequency Subbands Feature Extraction and Multibranch One-Dimension Convolutional Neural Network

Table 2

CNN branch HH.

Layer	Filters	Kernel size	Stride	Output size	Padding
Convolution; ReLU	16	16 × 1	2 × 1	256 × 16	Same
Convolution; Dropout (0.3)	24	4 × 1	1 × 1	256 × 24	Same
Average pooling		2 × 1	2 × 1	128 × 24
Max pooling		8 × 1	1 × 1	121 × 24	Valid
Convolution	32	3 × 1	1 × 1	121 × 32	Same
Max pooling		2 × 1	2 × 1	60 × 32	Valid