Abstract: Focusing on feature extraction in condition monitoring and fault diagnosis of mechanical equipment, a method based on the characteristics of white noise is presented to extract vibration mode from mechanical vibration signal. This method is a developed algorithm of empirical mode decomposition (EMD), which adaptively eliminates high frequency noise components and low frequency false components by applying the characteristics of normalized white noise under EMD, so the intrinsic mode set reflecting actual physical meaning of vibration signal is obtained. Hilbert transform is performed to the extracted intrinsic mode set, and the Hilbert time-frequency feature of observed signals are extracted. In the whole feature extracting process, the construction of general basis function described by some parameters and related filter function is unnecessary, and any prior information about the observed signal is no more required, so the method has a better applicability in actual applications. Both computer simulation and rotor set experimental results verify this approach is feasible and effective.

Key words: Feature extraction, Hilbert-Huang transform, Statistical characteristics of white noise, Vibration mode