An improved EEMD method based on the adjustable cubic trigonometric cardinal spline interpolation

The empirical mode decomposition (EMD) has recently emerged as an efficient tool to adaptively decompose non-stationary signals for nonlinear systems, which has a wide range of applications such as automatic control, mechanical engineering and medicine and biology. A noise-assisted variant of EMD named ensemble empirical mode decomposition (EEMD) have been proposed to alleviate the mode mixing phenomenon. In this paper, we proposed an improved EEMD method, namely cardinal spline interpolation based EEMD (C-EEMD), by optimizing the sifting procedure. Specifically, we employ the adjustable cubic trigonometric cardinal spline interpolation (CTCSI) to accurately represent free curves, other than the original one used in the traditional EEMD. The new interpolation approach can be used to build the mean curve in a more precise way. By virtue of CTCSI, we can therefore obtain the mean value curve from midpoints of the local maxima and minima by just one interpolation operations, which saves almost half the computational cost. Extensive experimental results on synthetic data and real EMI signals clearly demonstrate the superiority of the proposed method, compared to the state-of-the-arts.