Oumar I. Traoré, Nathalie
Favretto-Cristini, Laurent Pantera, Paul Cristini, S. Viguier-Pla,
Philippe Vieu (2017)
Which Methods and Strategies to Cope with Noise Complexity for an
Effective Interpretation of Acoustic Emission Signals in Noisy
Nuclear Environment?
Acta Acustica,
ISSN:
- Classification
- 62H12; 62H25; 62M15; 62P12
- Keywords
acoustic emission,
structural changes detection, signal denoising, nuclear environment
Abstract : Prior to an effective interpretation of AE signals
in noisy nuclear environment, it is necessary to define an appropriate
strategy for detecting structural changes and denoising signals,
according to the stochastic behavior of the noise and its level. The
ability and efficiency of various methods are studied here. One
conclusion of this work is that there is no absolute method which is
efficient regardless the type of noise and its level. In a high SNR
context, the continuous wavelet transform method is the best tool for
analysis of the signal structure, whatever the type of noise. For
denoising problem, the spectral subtraction and the singular spectrum
analysis methods are robust to changes in the stochastic behavior of
noise. However, the spectral subtraction has a higher SNR improvment
potential. We suggest that each selected method can be combined with
the other ones in order to further improve the denoising efficiency
and to get excellent results regardless the type of noise. The impact
of the different selected denoising methods on some classical AE
parameters shows that, whatever the type of noise and the denoising
method, AE parameters associated with the waveform are better restored
that those associated with the acoustic activity.