Limits...
A High-Resolution Demodulation Algorithm for FBG-FP Static-Strain Sensors Based on the Hilbert Transform and Cross Third-Order Cumulant.

Huang W, Zhen T, Zhang W, Zhang F, Li F - Sensors (Basel) (2015)

Bottom Line: The Hilbert transform is proposed for changing the Gaussian distribution of the two FBG-FPs' reflection spectra, and a cross third-order cumulant is used to use the results of the Hilbert transform and get a group of noise-vanished signals which can be used to accurately calculate the wavelength difference of the two FBG-FPs.The benefit by these processes is that Gaussian noise in the spectra can be suppressed completely in theory and a higher resolution can be reached.As a result, a static-strain resolution of 0.9 nε under laboratory environment condition is achieved, showing a higher resolution than the traditional cross-correlation method.

View Article: PubMed Central - PubMed

Affiliation: Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China. hwzhu@semi.ac.cn.

ABSTRACT
Static strain can be detected by measuring a cross-correlation of reflection spectra from two fiber Bragg gratings (FBGs). However, the static-strain measurement resolution is limited by the dominant Gaussian noise source when using this traditional method. This paper presents a novel static-strain demodulation algorithm for FBG-based Fabry-Perot interferometers (FBG-FPs). The Hilbert transform is proposed for changing the Gaussian distribution of the two FBG-FPs' reflection spectra, and a cross third-order cumulant is used to use the results of the Hilbert transform and get a group of noise-vanished signals which can be used to accurately calculate the wavelength difference of the two FBG-FPs. The benefit by these processes is that Gaussian noise in the spectra can be suppressed completely in theory and a higher resolution can be reached. In order to verify the precision and flexibility of this algorithm, a detailed theory model and a simulation analysis are given, and an experiment is implemented. As a result, a static-strain resolution of 0.9 nε under laboratory environment condition is achieved, showing a higher resolution than the traditional cross-correlation method.

No MeSH data available.


Related in: MedlinePlus

The simulation average values τ with the two different methods at different SNR levels.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4482008&req=5

sensors-15-09928-f007: The simulation average values τ with the two different methods at different SNR levels.

Mentions: In addition, 50 pairs of sensing signals with random relative Gaussian noise SNR level from −5 dB to 12 dB are used for proving our conclusion. The traditional cross-correlation method and cross third-order cumulant method are used to deal with these data, respectively, and an average processing of cross-correlation results is made at each SNR level. The simulation average values τ with the two different methods are shown in Figure 7. We can find that the estimation value τ by the cross third-order cumulant method is more accurate with respect to the actual wavelength difference than the traditional cross-correlation method whether the SNR of the sensing signals is high or low. Especially, the traditional cross-correlation method fails to work well when the SNR is less than −3 dB, but the cross third-order cumulant method presents a relatively high precision in this case.


A High-Resolution Demodulation Algorithm for FBG-FP Static-Strain Sensors Based on the Hilbert Transform and Cross Third-Order Cumulant.

Huang W, Zhen T, Zhang W, Zhang F, Li F - Sensors (Basel) (2015)

The simulation average values τ with the two different methods at different SNR levels.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4482008&req=5

sensors-15-09928-f007: The simulation average values τ with the two different methods at different SNR levels.
Mentions: In addition, 50 pairs of sensing signals with random relative Gaussian noise SNR level from −5 dB to 12 dB are used for proving our conclusion. The traditional cross-correlation method and cross third-order cumulant method are used to deal with these data, respectively, and an average processing of cross-correlation results is made at each SNR level. The simulation average values τ with the two different methods are shown in Figure 7. We can find that the estimation value τ by the cross third-order cumulant method is more accurate with respect to the actual wavelength difference than the traditional cross-correlation method whether the SNR of the sensing signals is high or low. Especially, the traditional cross-correlation method fails to work well when the SNR is less than −3 dB, but the cross third-order cumulant method presents a relatively high precision in this case.

Bottom Line: The Hilbert transform is proposed for changing the Gaussian distribution of the two FBG-FPs' reflection spectra, and a cross third-order cumulant is used to use the results of the Hilbert transform and get a group of noise-vanished signals which can be used to accurately calculate the wavelength difference of the two FBG-FPs.The benefit by these processes is that Gaussian noise in the spectra can be suppressed completely in theory and a higher resolution can be reached.As a result, a static-strain resolution of 0.9 nε under laboratory environment condition is achieved, showing a higher resolution than the traditional cross-correlation method.

View Article: PubMed Central - PubMed

Affiliation: Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China. hwzhu@semi.ac.cn.

ABSTRACT
Static strain can be detected by measuring a cross-correlation of reflection spectra from two fiber Bragg gratings (FBGs). However, the static-strain measurement resolution is limited by the dominant Gaussian noise source when using this traditional method. This paper presents a novel static-strain demodulation algorithm for FBG-based Fabry-Perot interferometers (FBG-FPs). The Hilbert transform is proposed for changing the Gaussian distribution of the two FBG-FPs' reflection spectra, and a cross third-order cumulant is used to use the results of the Hilbert transform and get a group of noise-vanished signals which can be used to accurately calculate the wavelength difference of the two FBG-FPs. The benefit by these processes is that Gaussian noise in the spectra can be suppressed completely in theory and a higher resolution can be reached. In order to verify the precision and flexibility of this algorithm, a detailed theory model and a simulation analysis are given, and an experiment is implemented. As a result, a static-strain resolution of 0.9 nε under laboratory environment condition is achieved, showing a higher resolution than the traditional cross-correlation method.

No MeSH data available.


Related in: MedlinePlus