Rolling Bearing Fault Detection Based on the Teager Energy Operator and Elman Neural Network

Table 4

Training data.

Signal type; (1750 rpm)	Target output	Input of the Elman neural network
Normal	(1,0,0)	0.0005	0.0005	0.1769
	(1,0,0)	0.0006	0.0005	0.1760
	(1,0,0)	0.0007	0.0005	0.1757
	(1,0,0)	0.0007	0.0005	0.1765
	(1,0,0)	0.0007	0.0006	0.1760
Inner ring fault	(0,1,0)	0.3274	0.8753	0.1548
	(0,1,0)	0.3269	0.8751	0.1548
	(0,1,0)	0.3267	0.8749	0.1548
	(0,1,0)	0.3269	0.8750	0.1548
	(0,1,0)	0.3192	0.8833	0.1551
Outer ring fault	(0,0,1)	4.6878	0.8272	0.0577
	(0,0,1)	4.6851	0.8230	0.0576
	(0,0,1)	4.6846	0.8216	0.0576
	(0,0,1)	4.6821	0.8199	0.0576
	(0,0,1)	4.6840	0.8188	0.0576