A new processing method for the end effect problem of empirical mode decomposition

In the empirical mode decomposition (EMD) method, the problem of end effect exists. By the mirror extending approach, the uniformity of the sifting process of EMD is implemented, but the problem of end pollution still exists. In the paper, an adaptive sifting approach is proposed. In the approach, the end extrema are determined by data itself, so, the Envelopes of data are calculated more reasonably. The approach not only implements the uniform of the sifting process of EMD, but also eliminates the end pollution, so that EMD method can be used more effectively.