A mean value mode decomposition method based on local interpolation by rational spline and its applications in fault diagnosis

The overshoots and undershoots occurred In the process of screening and the influence of the envelope curve shape because of end effect in EMD were discussed in this paper. In order to overcome the disadvantages in EMD mentioned above, a new signal decomposition method which is got mean value by an Weighted average of three adjacent nodes at first and obtain the average envelope based on local interpolation used the rational spline basis functions is proposed in this paper. Numerical simulation with a non-stationary, nonlinear data series has proved that the new decomposition method can avoid the error due to overshoots and undershoots during signal decomposition effectively, and it can restraining the end effect Better than original EMD. Finally, the new mean-value mode decomposition method based on local interpolation by rational spline can get accurate fault feature_because of misalignment from the radial vibration signal in a rotor system and therefore it can be applied successfully in the field of mechanical failure diagnosis.