文档介绍：Mechanism and Machine Theory 44 (2009) 697–709
Contents lists available at ScienceDirect
Mechanism and Machine Theory
journal homepage: ate/mechmt
A novel method for multi-fault diagnosis of rotor system
Jing JianPing *, Meng Guang
State Key Laboratory of Mechanical System and Vibration, Shanghai Jiaotong University, Shanghai 200240, PR China
article info abstract
Article history: The diagnosis for multiple faults is still a big challenge for monitoring and maintenance of
Received 5 October 2006 rotating machinery. In this study, the blind source separation (BSS) technique is proposed
Received in revised form 1 May 2008 to separate the vibrational features produced by several faults existing in a rotor. A statistic
Accepted 4 May 2008
variable-based algorithm, decorrelation method, is employed to analyze the signals of the
Available online 20 June 2008
rotors with multi-faults, which are simulated by numerical and experimental methods. The
separation results show that the typical ponents of the frequency spectrum of
each fault are separated essfully. The investigation demonstrates that it is feasible and
Keywords:
helpful to adopt BSS technique to diagnose the multi-faults of a rotor.
Multiple faults
Blind source separation Ó 2008 Elsevier Ltd. All rights reserved.
Rotor
Vibration
0. Introduction
For the fault diagnosis of rotating machinery, there have been several methods and theories such as spectrum method,
orbit method, and wavelet method [1,2] available in fault identification. These methods have been used in the practical diag-
nosis of rotor faults and proven to be effective in identifying most of mon faults such as unbalance, oil-whirl, rub and
impact-shaft misalignment. However, when multiple faults exist, vibrations excited by several faults bined with each
other linearly or nonlinearly. This makes the observed vibration signals plex and difficult to identify each fault in
terms of traditional methods. Now, the multi-fa