Fault Diagnosis of Rolling Element Bearing Based on Improved Ensemble Empirical Mode Decomposition

Abstract

The fault signal of rolling bearing has the characteristic of non-stationary, nonlinear and so on, the mode mixing phenomenon may occur in the process of empirical mode de-composition. Ensemble empirical mode decomposition (EEMD) algorithm is introduce random Gaussian white noise sequence in the original signal to change the local time of the signal span, Which can inhibit the mode mixing phenomenon in the process of the conventional empirical mode decomposition. On the basis of the principles of the EEMD, This paper introduced the Amplitude standard deviation criterion to select the EEMD parameters. And for each intrinsic mode function (IMF) components decomposed by correlation coefficient method to extract the effect intrinsic mode component, then through threshold and reconstructing each effective intrinsic mode function. Finally the envelop spectrum of the signal was analyzed, extracted the fault characteristics of the rolling bearings. Simulation and fault signals experimental results show that, EEMD method can be effectively applied to fault diagnosis of rolling bearings.