Research on EMD method based on reproducing kernels

Empirical Mode Decomposition (EMD), the core of Hilbert-Huang transformation, is reckoned on the algorithm of the extrama extending and generation of envelope curves. A novel EMD algorithm was proposed based on reproducing kernels with good local property such as odd order vanishing moments, rapid attenuation of scaling, symmetry and regularity. At last, simulations on two non-stationary signals show the novel EMD algorithm is suitable to analyze the nonlinear and non-stationary signal.