Hits:

Abstract:Nonstationary signal generally exhibits multifractal structure, different from a simple monofractal structure that can be depicted by a single scaling exponent, which requires multiple scaling exponents for a full description of the dynamic behavior of signals. In this case, multifractal detrended fluctuation analysis (MF-DFA) is developed for the multifractal characteristics analysis of nonstationary signals. Firstly, we using MF-DFA method to process and analyze several typical signals and then apply it to investigate heart rate variability signals. The results indicate the MF-DFA method is a reliable tool of detecting the monofractality and mulifractality of time series. Furthermore, MF-DFA method can effectively distinguish the different heart rate variability signals. The study shows that MF-DFA method is a promising technique of detection and determination of mulifractality of nonstationary time series. © 2015 IEEE.

Translation or Not:no