Shaft Orbit Feature Based Rotator Early Unbalance Fault Identification

Abstract Feature extraction is crucial to rotating machinery prognosis, which is an important aspect of condition monitoring as well as maintenance program, since the quality of feature will impact the result significantly. Vibration signals are commonly used as the source for feature extraction during the prognosis process, especially the energy feature of fundamental frequency (which is written as 1X), 2X, 3X, 1/2X, etc. Yet this kind of feature shows insufficiency for identifying stages of performance degradation and classifying the type of early fault, therefore researchers focused mainly on improving the methods of feature extraction to solve this problem. However, features extracted from vibration signals always ignore some fault information such as kinematics information and phase information, thus other source of feature is needed to provide supplement or even substitute for higher efficiency and sharpness of separation in rotating machinery prognosis, which are strongly demanded by today's complex and advanced machines. This paper introduced one kind of classic feature source: shaft orbit, which is widely used in traditional diagnosis for failure classification, into prognosis, and its effectiveness is verified in rotor early unbalance fault identification using features extracted from it, compared with energy features of frequency band extracted from vibration signals. Result shows that shaft orbit feature can be used in identifying different early fault stages of rotor unbalance, which indicates that utilizing shaft orbit as source of feature extraction can provide a new approach of getting early fault features in rotating machinery prognosis.

[1]  Junyan Yang,et al.  Intelligent fault diagnosis of rolling element bearing based on SVMs and fractal dimension , 2007 .

[2]  Gang Wang,et al.  Diagnosis of bearing incipient faults using fuzzy logic based methodology , 2010, 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery.

[3]  Weiji Wang,et al.  Purification and feature extraction of shaft orbits for diagnosing large rotating machinery , 2005 .

[4]  Jyoti K. Sinha,et al.  Condition monitoring of rotating machines using on-shaft vibration measurement , 2012 .

[5]  Zhike Peng,et al.  Identification of the shaft orbit for rotating machines using wavelet modulus maxima , 2002 .

[6]  Sheng-Fa Yuan,et al.  Fault diagnostics based on particle swarm optimisation and support vector machines , 2007 .

[7]  Lei Lu,et al.  Incipient Bearing Fault Diagnosis Based on Improved Hilbert-Huang Transform and Support Vector Machine , 2011 .

[8]  Joseph Mathew,et al.  Rotating machinery prognostics. State of the art, challenges and opportunities , 2009 .

[9]  Jay Lee,et al.  Degradation Assessment and Fault Modes Classification Using Logistic Regression , 2005 .

[10]  Haiyi Zhu,et al.  Assessment of Rotor Degradation in Steam Turbine Using Support Vector Machine , 2009, 2009 Asia-Pacific Power and Energy Engineering Conference.

[11]  Liangsheng Qu,et al.  Orbit Complexity: A New Criterion for Evaluating the Dynamic Quality of Rotor Systems , 1993 .

[12]  Daming Lin,et al.  A review on machinery diagnostics and prognostics implementing condition-based maintenance , 2006 .

[13]  Jianzhong Zhou,et al.  Fault diagnosis based on Walsh transform and support vector machine , 2008 .

[14]  Zacharias Voulgaris,et al.  A Novel Automated Feature Extraction Method for Fault Diagnosis of Rotating Mechanical Systems , 2010 .