Vibration signal denoising method for drive roller bearing of mine-used belt conveyor

Abstract

EMD has modal aliasing and ICA requires the number of collected vibration signal should not be less than the number of source signal. In view of the above problems that existed in existing vibration signal denoising method, a vibration signal denoising method for drive roller bearing of mine-used belt conveyor was proposed which was based on EEMD and FastICA. Firstly, collected vibration signal is decomposed by EEMD algorithm, so as to obtain several IMF components at different scales that containing faults characteristic frequency. Then, the IMF components are reconstructed based on correlation coefficient to obtain characteristic signal and virtual noise signal, which are formed into input matrix as input of FastICA algorithm. Finally, FastICA algorithm is used to separate signal and noise to achieve purpose of signal denoising. The experimental results verify feasibility and effectiveness of the method.

Keywords

• mine-used belt conveyor
• drive roller bearing
• fault diagnosis of roller bearing
• vibration signal denoising
• ensemble empirical mode decomposition
• fast independent component analysis