期刊
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
卷 56, 期 11, 页码 4557-4570出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIE.2009.2026211
关键词
Fault diagnosis; frequency estimation; Hilbert transform (HT); induction motors; time-frequency analysis; wavelet transform (WT)
资金
- European Community's Seventh Framework Program [FP7/2007-2013]
- Research Project Power Plant Robustification based on fault Detection and Isolation (PRODI) algorithms [224233]
In this paper, a new method for detecting the presence of broken rotor bars is presented. The proposed approach is valid for induction machines started at constant frequency and consists of extracting the instantaneous frequency (IF) of the left sideband harmonic (LSH) from the start-up current (LSHst), via the Hilbert transform. It is shown that, in the case of machines with one or several broken bars, the IF of the LSHst exhibits a very characteristic and easy to identify pattern, which is physically justified. This paper also shows that, if the IF of the LSHst is represented against the slip, a universal fault indicator (nondependent neither on the machine characteristics nor on the starting conditions) can be defined. This fault indicator consists of the correlation between the experimental IF of the LSHst and its theoretical evolution. This approach is theoretically introduced and experimentally validated by testing a commercial motor in faulty and healthy conditions, under different operating conditions.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据