DocumentCode :
1945065
Title :
Feature extraction of frictional vibration based on CWT time-frequency image
Author :
Huang, Chaoming ; Yu, Hongliang ; Guan, Delin ; Li, Guobin ; Wu, Qili
Author_Institution :
Coll. of Maritime Eng., Dalian Maritime Univ., Dalian, China
fYear :
2010
fDate :
13-15 Aug. 2010
Firstpage :
192
Lastpage :
196
Abstract :
A method to extract features of frictional vibration signals was developed, using continuous wavelet transform (CWT) time-frequency image (TFI) and image processing to analyze the vibration signals, since the frictional vibration signals contain a wealth of information about the interface´s fault state and friction wear state. TFIs were generated by CWT theory from the vibration signals produced by cylinder and piston ring samples sampled from 6S50MC marine diesel engine under two different lubrication statuses. And characteristic parameters obtained by image segmentation of TFIs were analyzed in order to establish the intrinsic relationship between the characteristic parameters and friction wear status. The results show that regular variations occur to the characteristic parameters with the conduct of the friction process, and characteristic parameters of different lubrication condition show significant differences. The CWT TFI characteristic parameters of vibration signals can be used to quantitatively describe vibration signal characteristic.
Keywords :
diesel engines; feature extraction; friction; image segmentation; lubrication; pistons; shapes (structures); structural engineering computing; time-frequency analysis; vibrations; wavelet transforms; wear; 6S50MC marine diesel engine; continuous wavelet transform time-frequency image; cylinder; feature extraction; friction wear state; friction wear status; frictional vibration signal; image processing; image segmentation; interface fault state; lubrication status; piston ring samples; Continuous wavelet transforms; Fault diagnosis; Fractals; Friction; Image segmentation; Time frequency analysis; Vibrations;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Control and Information Processing (ICICIP), 2010 International Conference on
Conference_Location :
Dalian
Print_ISBN :
978-1-4244-7047-1
Type :
conf
DOI :
10.1109/ICICIP.2010.5564312
Filename :
5564312
Link To Document :
بازگشت