ISSN: 0090-3973
Published Online: 13 April 2007
Page Count: 8

Fault Classification of Water Hydraulic System by Vibration Analysis with Support Vector Machine

Chen, H. X.
School of Mechanical and Aerospace Engineering, Nanyang Technological University,

Chua, Patrick S.K.
School of Mechanical and Aerospace Engineering, Nanyang Technological University,

Lim, G. H.
School of Mechanical and Aerospace Engineering, Nanyang Technological University,

(Received 21 July 2006; accepted 23 January 2007)

Abstract

This paper presents a new neural network approach to the fault diagnosis of a water hydraulic system based on the wavelet analysis of a vibration signal. A novel feature of this approach is that the vibration signals acquired from the water hydraulic motor are employed for analysis. Wavelet transform (WT) is first applied as a feature extraction technique to analyze the time-domain vibration signal. The performance of support vector machine (svm) is then investigated and compared with the conventional neural network. The results confirm the applicability of the proposed method for the fault detection in a modern water hydraulic system.

Keywords:
fault diagnosis, support vector machine, water hydraulic system, feature extraction, wavelet transform, neural network

Paper ID: JTE100734
DOI: 10.1520/JTE100734
ASTM International is a member of CrossRef.