Fault Early Warning Based on Improved Deep Neural Network of Auto-Encoder

Table 2

Parameter of the model.

Layer	Structure name	Structure parameter	Number of channels	Output size
1	Input	(3600,1)	1	(3600,1)
2	Convolution layer	(1,1)	5	(3600,1)
3	Convolution layer	(3,1)	4	(1200,1)
4	Convolution layer	(3,1)	8	(400,1)
5	Convolution layer	(3,1)	4	(134,1)
6	Convolution layer	(3,1)	8	(400,1)
7	Convolution layer	(3,1)	4	(1200,1)
8	Convolution layer	(1,1)	5	(3600,1)
9	Convolution layer	(3600,1)	1	(3600,1)