Transient Signal Analysis Based on Hilbert-Huang Transform

This paper investigates the utility of Hilbert-Huang transform (HHT) for transient signal analysis. HHT represents a one-dimensional time series in a 2D time-frequency domain, thus is suited for analyzing nonstationary events commonly seen in rotary mechanical and electromechanical systems, where inception and propagation of structural defects during machine operations cause local perturbation to the periodic structural vibration signals. Using a synthetic signal, the analytical framework for HHT is first introduced. Simulation results are then experimentally evaluated based on vibration signals measured from a realistic test bed for rolling bearings. The results demonstrate that HHT provides an effective signal decomposition tool for machine health monitoring