A Multi-Wave-Packets Decomposition Algorithm of Vibration Signal
From the perspective of mechanic wave propagation, a new approach to decompose a vibration signal is proposed in this paper. Define a series wave-packet mode functions, think a vibration signal of the sensor correcting is many resonance wave-packets linear super-position after propagation, reflectio...
Saved in:
Published in | Noise & vibration worldwide Vol. 42; no. 10; pp. 3 - 8 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
London, England
SAGE Publications
01.11.2011
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | From the perspective of mechanic wave propagation, a new approach to decompose a vibration signal is proposed in this paper. Define a series wave-packet mode functions, think a vibration signal of the sensor correcting is many resonance wave-packets linear super-position after propagation, reflection, diffraction and dispersion. The multi-wave-packets mode decomposition algorithm of vibration signal mainly includes two parts, that is, wave-packet center detection and extraction of wave-packet parameters. Assume the resonance wave-packet center signal value is maximum in the local time-frame, each possible wave-packet center mode fitting is done with an additional-window sliding manner, at the same time, corresponding to parameters are extracted. These parameters represent every mechanical fluctuation of the sensor collecting. Meanwhile, they can recover the signal better and provide many characteristics for signal detection and fault diagnosis. By applying the algorithm to the In-line vibration inspection of anchor screws in light rail, it has been proven that the wave-packet mode parameters obtained from the algorithm is very useful for connection looseness testing and fault diagnosis. |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 0957-4565 2048-4062 |
DOI: | 10.1260/0957-4565.42.10.3 |