A novel denoising method based on ensemble empirical mode decomposition principle component analysis

Yang Feng, Xiao Diao Huang, Jie Chen, Rong Jing Hong

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Equipment performance degradation model is a key component in its diagnosis and prognosis models. As to slewing bearing, vibration signals are usually non-stationary and with strong white noise, which makes it very difficult to extract useful information from the signals. Therefore, an Ensemble Empirical Mode Decomposition – Principle Component Analysis (EEMD - PCA) based method was proposed to denoise the vibration signals and performance degradation model was established by PCA. To verify the proposed method, an experiment was conducted and the life cycle vibration signals were acquired. After denoising, performance degradation model was established to explain the denoising performance. Results show that the proposed method acquired a better denoising performance than EEMD-MSPCA, which provides a potential for further research.

Original languageEnglish
Title of host publicationMechanical Components and Control Engineering III
EditorsWeimin Ge
PublisherTrans Tech Publications Ltd
Pages1157-1261
Number of pages105
ISBN (Electronic)9783038353126
DOIs
StatePublished - 2014
Event3rd Asian Pacific Conference on Mechanical Components and Control Engineering, ICMCCE 2014 - Zhuhai, China
Duration: 20 Sep 201421 Sep 2014

Publication series

NameApplied Mechanics and Materials
Volume668-669
ISSN (Print)1660-9336
ISSN (Electronic)1662-7482

Conference

Conference3rd Asian Pacific Conference on Mechanical Components and Control Engineering, ICMCCE 2014
Country/TerritoryChina
CityZhuhai
Period20/09/1421/09/14

Keywords

  • EEMD-PCA
  • Non-stationary signal
  • Performance degradation model
  • Slewing bearing

Fingerprint

Dive into the research topics of 'A novel denoising method based on ensemble empirical mode decomposition principle component analysis'. Together they form a unique fingerprint.

Cite this