The temperature or/and strain measurement accuracy, resolution and response of distributed optical fiber sensors based on Rayleigh, Raman and Brillouin scattering are mainly limited by their weak signals and low signal-noise ratio (SNR). To improve the SNR and response characteristic simultaneity, the wavelet transform de-noising technology was introduced to distributed optical fiber sensor system. The mother wavelet and the processing of thresholds were optimized in order to get better SNR and minimize the impact to signal response. A Raman scattering distributed optical fiber sensing experimental system was set and the signal process arithmetic was realized. Experimental tests indicate that wavelet de-noising method can notably improve SNR and guarantee the spatial resolution simultaneously. Final contrast tests suggest that not only the temperature resolution was improved by more than 50%, but also the measurement accuracy and period were improved greatly.
[1]
郭宁,et al.
Distributed optical fiber temperature sensor system
,
2001
.
[2]
I. Johnstone,et al.
Ideal spatial adaptation by wavelet shrinkage
,
1994
.
[3]
Hong-Ye Gao,et al.
Wavelet Shrinkage Denoising Using the Non-Negative Garrote
,
1998
.
[4]
Nirmal K. Bose,et al.
Properties determining choice of mother wavelet
,
2005
.
[5]
Stéphane Mallat,et al.
A Theory for Multiresolution Signal Decomposition: The Wavelet Representation
,
1989,
IEEE Trans. Pattern Anal. Mach. Intell..
[6]
I. Johnstone,et al.
Adapting to Unknown Smoothness via Wavelet Shrinkage
,
1995
.
[7]
Branko Glisic,et al.
Development of distributed strain and temperature sensing cables
,
2005,
International Conference on Optical Fibre Sensors.