Research on the intrinsic mode function (IMF) criterion in EMD method

Empirical mode decomposition (EMD) is a self-adaptive signal-processing method, which has been applied in non-stationary signal-processing successfully. Aiming at the problem of intrinsic mode function (IMF) criterion in the EMD method, the energy difference tracking method is proposed in this paper according to the integrity and orthogonality of the IMFs and used to define the IMF in the sifting process. By analyzing the simulated and actual signals, it is confirmed that the IMFs defined by the energy difference tracking method meet the orthogonality condition and reflect the intrinsic and real information of the analysed signal. By comparing the energy difference tracking method with the Cauchy-type convergence criterion, it is demonstrated that the IMFs obtained by the energy difference tracking method can reflect the intrinsic information included in the signal more clearly and their index of orthogonal (IO) is smaller.