An Accurate Prediction Algorithm of RUL for Bearings: Time-Frequency Analysis Based on MRCNN

Table 2

Comparison results of this research and related researches in the PRONOSTIA dataset.

Testing dataset	Current time (s)	Actual RUL (s)	Predicted RUL (s)		[24]	[5]	[25]
Bearing1_3	18010	5730	5470	4.53	37	-0.35	7.62
Bearing1_4	11380	2890	3360	-16.26	80	5.60	-157.71
Bearing1_5	23010	1610	1970	-22.36	9	100.00	-72.57
Bearing1_6	23010	1460	1290	11.64	-5	28.08	0.93
Bearing1_7	15010	7570	6870	9.24	-2	19.55	85.99
				12.81	26.6	30.71	64.96
Score				0.47	0.46	0.45	0.35