Crossing Sifting Method Based on Huang Transform

Crossing sifting method is developed based on Huang transform, which is used for analyzing nonlinear and non-stationary time series. The novel method can obtain the approximate mean of signal upper and lower envelopes directly by interpolating the crossing series using cubic spline function. After finding the local extrema of time series, the crossing series can be located by neighbor extrema. The empirical mode method of Huang Transform and new method are compared through several examples. The simulation results show the new method is a simple and valid one, and it can decompose a signal into some satisfying intrinsic mode functions.