An Improved EMD Algorithm for Self-Adaptive Adjustment of Reference Point Position

Abstract

Empirical Mode decomposition (EMD), invented by Huang et al., National Aeronautics and Space Administration (NASA), is an advanced signal processing method, which can effectively obtain the time-frequency characteristics of non-stationary signals. However, when the upper and lower envelope of the signal is used to construct the cubic spline curve to separate the signals of different modes, the signal distortion problems such as end effect will appear. On the basis of studying the existing methods of this problem, an adaptive method of extremum point position is proposed in this paper, which can reduce the distortion degree in the extraction process of different mode signals. The method takes full account of the intrinsic characteristics of signals, and the effect is better in a variety of evaluation criteria.