Application of instantaneous phase analysis based on empirical mode decomposition

A novel method of instantaneous phase analysis based on empirical mode decomposition is presented. Firstly, the empirical mode decompositions of the vibration signals result in the intrinsic mode functions with different time scales. Then, the Hilbert transform is applied to each intrinsic mode function, therefore, the instantaneous phase is obtained. Finally, the fast Fourier transform is applied to the instantaneous phase, and the feature is extracted by the frequency spectrum analysis. The method is introduced in theory and is used to diagnose bearing faults. In order to recognize the fault features, the instantaneous phase analysis and the fast Fourier transform are applied to the selected intrinsic mode functions standing for the bearing faults. The experimental results show that the instantaneous phase analysis based on empirical mode decomposition can effectively diagnose the bearing faults.