Affiliations: College of Mechanical and Electrical Engineering,
Central South University, Changsha 410083, China | Hunan Provincial Key Laboratory of Health Maintenance
for Mechanical Equipment, Hunan University of Science and Technology, Xiangtan
411201, China
Abstract: This paper proposes a new approach combining autoregressive (AR)
model and fuzzy cluster analysis for bearing fault diagnosis and degradation
assessment. AR model is an effective approach to extract the fault feature, and
is generally applied to stationary signals. However, the fault vibration
signals of a roller bearing are non-stationary and non-Gaussian. Aiming at this
problem, the set of parameters of the AR model is estimated based on
higher-order cumulants. Consequently, the AR parameters are taken as the
feature vectors, and fuzzy cluster analysis is applied to perform
classification and pattern recognition. Experiments analysis results show that
the proposed method can be used to identify various types and severities of
fault bearings. This study is significant for non-stationary and non-Gaussian
signal analysis, fault diagnosis and degradation assessment.
