A new method for mitigation of end effect in empirical mode decomposition

Empirical mode decomposition(EMD) is a time-frequency analysis method for non-stationary and nonlinear signal. In the iteration process of EMD, at least one data point outside each end is required to build the spline. However, the extending data points which are not chosen properly will cause vibration and distortion phenomenon in EMD. In order to mitigate that end effect, a novel extrema points extending method is proposed: End mirror extending is used in high frequency, while least square polynomial extending is used in low frequency. Numerical result proves that the proposed method can mitigate the end effect in EMD effectively with minor decomposition error.