A Study of the Equipment Fault Diagnosing Method Based on EMD and Holo-spectrum

We first introduce the empirical mode decomposition （EMD） and holo-spectrum analysis. Then we decompose the vibration signal of the equipment by EMD, re-organize the signal closely associated with the original signal, eliminate the noise signal better, and analyze the holographic re-signal spectrum, which makes the amplitude, frequency and phase combinecompletely, reflecting the equipment vibration morphology fully. Particularly applying a two-dimensional holo-spectrum can obtain the direction of rotation order harmonic, size, shape, and the relation- . ship between order harmonic generation rotor vibrations, and so on. The traditional signal processing methods can not eliminate noise well. We apply this method to the rotor vibration signal of fault diagnosis, eliminate noise and reflect the equipment vibration characteristics well.