Diagnosis of rotating machine defects by vibration analysis

This work presents the analysis of vibration signals a ball bearing, available on platform Case Western Reserve University in form MATLAB files. By proposing an approach for diagnosis any rotating machine and detects failed components simple way, proposed consists several methods steps each complementary to other. First decomposing signal by feature mode decomposition (FMD) method then, according Kurtosis values selecting useful after selection computing sum called residual original signal. The simplicity makes through application band pass filtering Hilbert transform, but peaks represents frequency components.