An extended EEMD method for localized faults detection of a planetary gearbox

Jing Liu, Linfeng Wang, Hanjie Tan, Liming Wang, Zaigang Chen, Yimin Shao

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

An accurate incipient localized faults diagnosis method is very helpful for preventing serious accidents of planetary gearboxes. Although useful, the ensemble empirical mode decomposition (EEMD) method still has some issues in selecting the effective intrinsic mode functions (IMFs). To overcome these issues, an improved encircled energy EEMD method (EE-EEMD) combined with the EEMD algorithm, mirror extending method, Teager energy operator demodulation method, and EE index selection method is presented to detect localized faults in the planet bearings, ring gear, planet gear, and sun gear of planetary gearboxes. The mirror extending method is applied to address the end extending issue of the EEMD method. The EE index is utilized to determine the effective IMFs from the EEMD method. The energy separation algorithm is used to calculate the instantaneous frequencies of the effective IMFs. The results from the EE index, kurtosis, and weight kurtosis selection methods are compared. The vibration signal models of the localized faults in the ring gear, planet gear, sun gear, planet bearing races, and planet bearing roller are used to illustrate the validity of the presented EE-EEMD method. An experimental investigation for a planetary gearbox considering a spall fault in the sun gear is conducted to validate the presented EE-EEMD method. It seems the presented EE-EEMD method can be utilized to detect localized faults in the components of planetary gearboxes.

Original languageEnglish
Pages (from-to)758-774
Number of pages17
JournalJournal of Testing and Evaluation
Volume47
Issue number2
DOIs
StatePublished - 13 Feb 2019
Externally publishedYes

Keywords

  • Diagnosis method
  • Encircled energy ensemble empirical mode decomposition
  • Localized faults
  • Planetary gearbox
  • Vibration

Fingerprint

Dive into the research topics of 'An extended EEMD method for localized faults detection of a planetary gearbox'. Together they form a unique fingerprint.

Cite this