Method for Bearing Fault Diagnosis Based on MSD-ICA-EA
Zongtian Liu · Bearing Jurnal Penelitian dan Kajian Teknik Sipil · 2008
Bearing faults are caused by several reasons and the original fault source signals are statistically correlative each other in practical problem,conventional independent component analysis is restricted.A method,integration of multiresolution subband decomposition and independent component analysis,is proposed.High frequency subband signal with amplitude modulation,separated by wavelet transform and fast ICA,is further analyzed by envelope analysis to diagnose bearing faults.Experimental result shows that MSD-ICA-EA is effective to solve the problem.