The Optimization Parameters of EEMD Applied to Rolling Bearing Fault Diagnosis

In order to restrain the phenomenon of the mode mixing in empirical mode decomposition (EMD) and improve decomposition precision, the ensemble empirical mode decomposition (EEMD) algorithm was introduced. But the effect of this algorithm depends on two important parameters, the amplitude of the added white noise and the number of ensemble trials. The energy standard deviation method is proposed, a rule of setting parameters in EEMD algorithm by investigating the effect of the added white noise to the decomposition results. Thus, the number of ensemble trials will be obtained under the given error. And this is a basis of obtaining adaptively EEMD algorithm. The resisted mode mixing performance of the EEMD algorithm and the effectiveness of the added white noise standard are verified by simulation test and the actual signals of rolling bearing with the inner raceway fault.