Enhanced Detection of Characteristic Vibration Signal of Generator Based on Self-Adapted Multi-Scale Top-Hat Transformation

Author(s): Yuling He*, Yuyang Zhang, Yuchao Meng, Guiji Tang, Weiqi Deng, Hao Zhong

Journal Name: Recent Advances in Electrical & Electronic Engineering
Formerly Recent Patents on Electrical & Electronic Engineering

Volume 11 , Issue 4 , 2018

Become EABM
Become Reviewer
Call for Editor

Graphical Abstract:


Objective: Aiming at the difficulty of extracting weak faulty vibration features of generator in practical performing, a novel method named self-adapted multi-scale top-hat transformation (SAMST) is proposed to enhance the detection of the characteristic vibration signals.

Methods: Different from other studies, this method employs the Sine-Structure Element (SSE) which is more similar to the signals appear in electrical systems to filter the noise. The most optimal scale of the SSE is obtained by using the method named Feature Amplitude Energy Radio (FAER) to enhance the characteristic faulty components. Experiment studies for an MJF-30-6 type fault simulating generator under the stator inter-turn short circuit fault confirm the advantages of this method.

Results: It is shown that the proposed method can not only enhance all of the three characteristic frequency components of the stator vibration signal respectively at 2f, 4f, and 6f (f is the electrical frequency), but also depress the strong background noises and retain the detailed information at the same time.

Conclusion: This method is probably to offer more convenience for the fault diagnosis of generator and therefore has practical application values.

Keywords: Characteristic vibration signal, enhanced detection, Feature Amplitude Energy Radio (FAER), generator, Self- Adapted Multi-Scale Top-Hat Transformation (SAMST), Sine-Structure Element (SSE).

Rights & PermissionsPrintExport Cite as

Article Details

Year: 2018
Published on: 12 February, 2018
Page: [418 - 424]
Pages: 7
DOI: 10.2174/2352096511666180213101258
Price: $25

Article Metrics

PDF: 6
PRC: 1