Back to Search Start Over

Fault diagnosis and location identification of rotor–stator rub-impact based on Hjorth parameters.

Authors :
Yu, Mingyue
Chen, Wangying
Lu, Yanjun
Source :
Engineering Failure Analysis. Aug2022, Vol. 138, pN.PAG-N.PAG. 1p.
Publication Year :
2022

Abstract

• Take advantage of sensibility of Activity parameter to vibration intensity to describe the intensity of signals and identify different rub-impact states. • Make use of the sensitivity of Mobility parameter to signal state to determine the occurrence of rub-impact fault. • The sensitivity of Complexity parameter to signal complexity will be relied on to identify the different rub-impact states of equipment. • Hjorth parameters can be treated as the characteristic parameters for identifying of rub-impact fault and location. Hjorth parameters reveals Activity, Mobility and Complexity of signals. From this, the paper has proposed the scheme to identify a rotor–stator rub-impact fault and occurrence location by Hjorth parameters. Firstly, the approach has given an analysis to the difference of Hjorth parameters of signals and verified the sensibility of Hjorth parameters to running state of equipment; secondly, in the same running state of equipment, the consistency and robustness of Hjorth parameters of signals from the same sensor have been analyzed and proved; thirdly, for verifying the sensitivity of sensors installation locations to rub-impact fault and occurrence locations, Hjorth parameters of signals from different sensors have been analyzed; finally, the calculated Hjorth parameters are used as feature vector, and K-nearest neighbor (KNN) classification algorithm is employed for identifying a rub-impact fault and determining the rub-impact locations. The results indicate that proposed method can accurately identify the rub-impact fault and occurrence locations. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
13506307
Volume :
138
Database :
Academic Search Index
Journal :
Engineering Failure Analysis
Publication Type :
Academic Journal
Accession number :
157388588
Full Text :
https://doi.org/10.1016/j.engfailanal.2022.106299