Abstract

Decomposing non-stationary signals using Empirical Mode Decomposition (EMD) highly facilitates signal analyses and processing. According to the original algorithm, EMD decomposes the input signal into useful Intrinsic Mode Functions (IMFs). However, EMD has some drawbacks. The first one is that the number of IMFs is not known in advance and can change with a small variation in the data. The second one is that EMD must be applied to a signal in its entirety, being not possible to proceed by parts either to reduce the computational load or to deal with real-time data. So, EMD it’s not feasible to use on long signals or for dealing with streaming data signals. These two drawbacks limit EMD practical application and are addressed in this text. A novel way to run EMD on any long or streaming signal is provided, while maintaining a constant number of IMF outputs. The method uses an innovative extension of the original EMD, called Flow Empirical Mode decomposition (FEMD), which applies EMD on sliding windows and ensures a fixed number of IMFs. Furthermore, it is demonstrated the successful use of FEMD on an Electrocardiogram (ECG) analyses. The developed FEMD software was made freely available.

Original languageEnglish
Title of host publicationIntelligent Systems and Applications
Subtitle of host publicationProceedings of the 2021 Intelligent Systems Conference (IntelliSys) Volume 3
EditorsKohei Arai
Place of PublicationCham
PublisherSpringer
Pages234-250
Number of pages17
Volume3
ISBN (Electronic)978-3-030-82199-9
ISBN (Print)978-3-030-82198-2
DOIs
Publication statusPublished - 2022
Event Intelligent Systems Conference, IntelliSys 2021 - Virtual, Online
Duration: 2 Sep 20213 Sep 2021

Publication series

NameLecture Notes in Networks and Systems
PublisherSpringer
Volume296
ISSN (Print)2367-3370
ISSN (Electronic)2367-3389

Conference

Conference Intelligent Systems Conference, IntelliSys 2021
CityVirtual, Online
Period2/09/213/09/21

Keywords

  • Bearings
  • ECG
  • Electrocardiogram
  • EMD
  • Empirical Mode Decomposition
  • FEMD
  • Sliding window

Fingerprint

Dive into the research topics of 'Flow Empirical Mode Decomposition'. Together they form a unique fingerprint.

Cite this