A Novel Multimode Fault Classification Method Based on Deep Learning

Table 2

Data description of in dataset for a given mode.

Dataset for different fault location	Dataset for different fault severity	Sample number; More/fewer	Fault type	Fault size /mm
		200/100	Normal (N)	0.000
		200/100	Inner race fault (IF)	0.007
		200/100	Inner race fault (IF)	0.014
		200/100	Inner race fault (IF)	0.021
		200/100	Outer race fault (OF)	0.007
		200/100	Outer race fault (OF)	0.014
		200/100	Outer race fault (OF)	0.021
		200/100	Roller fault (RF)	0.007
		200/100	Roller fault (RF)	0.014
		200/100	Roller fault (RF)	0.021