Loading...

Experimental investigation of empirical mode decomposition by reduction of end effect error

Momeni Massouleh, H ; Sharif University of Technology | 2019

370 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.physa.2019.122171
  3. Publisher: Elsevier B.V , 2019
  4. Abstract:
  5. Empirical mode decomposition as a complete data driven method is typically employed to obtain constitutive components of all kinds of signal including non-stationary and nonlinear. Extracting modes using this method is normally associated with multiple sources of errors such as stop criteria, end effects, interpolation function and etc. In order to reduce end effects errors in this paper a modified method is proposed based on combination of auto regressive model and mirror method. In this combination method, to extract some of first intrinsic mode functions of a given signal, auto regressive model is initially implemented to forecast tails of maximum and minimum envelops for only a short section of the signal at its both ends. Then the mirror method is employed to continue sifting process for remaining signal that has no enough extrema to employ auto regressive model. Advantages of the proposed combined method are analytically and experimentally assessed and comparisons with mirror method solutions are presented. © 2019 Elsevier B.V
  6. Keywords:
  7. Auto regressive model ; Empirical mode decomposition ; Experimental assessment ; Mirrors ; Signal processing ; Auto regressive models ; End effects ; Experimental assessment ; Mirror method ; Errors
  8. Source: Physica A: Statistical Mechanics and its Applications ; Volume 534 , 2019 ; 03784371 (ISSN)
  9. URL: https://www.sciencedirect.com/science/article/pii/S0378437119312592