Near-field multiple signal classification algorithm for acoustic emission source localization in rolling element rub-impact fault diagnostics

作者：Li Jing; Deng Aidong; Liu Dongying; Zhang Rui; Zhao Li

来源：ICIC Express Letters, 2016, 10(3): 663-668.

• Acoustic emission sources
• Multiple signal classification
• Multiple signal classification algorithm
• Multiple signal classification methods
• Near fields
• Propagation velocities
• Rub impact
• Wavelet packet transform(WPT)

Rub-impact source localization is an important indicator to evaluate incipient faults in rotating machine. The accurate evaluation of damage location can assist in early warning of failure. This paper proposes a Near-Field Multiple Signal Classification method (NFM) to estimate the location of acoustic emission (AE) source in rub-impact, using an array of sensors closed placed in a local region. This method combines the Wavelet Packet Transform (WPT) for the zero order extensional (S0) and flexural waves (A0) extraction, Modal Plate Wave Theory (MPWT) analysis for the revised propagation velocity and NFM algorithm to find the AE source position. The localization accuracy of rubbing source based on NFM, time of arrival (TOA), Delta-T and delay time beamforming (DTB) methods are compared by rubbing test carried on a test rig of rotation machinery. The results indicate that the improved method can localize rub fault more reliable and efficient than the others and is a helpful analysis tool for rub-impact fault diagnosis.

• 出版日期2016
• 单位东南大学