Hilbert Transform Approach to Central Wavelength Detection for Fiber Bragg Grating Sensors

Abstract

The accuracy and sensitivity of Fiber Bragg Grating sensors depends on signal processing approaches that detect the wavelength of the centeral peak in the reflection spectra. In the studies carried out so far, there are various noise that seriously affect the system, arising from the electronic elements in their structure and the environment in which they operate. In addition, depending on the coherence length and intensity of the light sources used, the effects such as unwanted interference in the reflection spectrum create noise. Therefore, the reflection spectrum of the FBG sensor is noisy. In recent years, filtering techniques and curve fitting methods etc. have become increasingly important to reduce the effect of this noise. In this study, it is revealed the Hilbert transform approach enables the detection of the more accurate central wavelength of the FBG sensor. This approach is very practical. Because the Hilbert transform already acts as a filter, this approach does not require a filter design, decomposition levels, or any other complex process as in other methods. To demonstrate that the proposed approach improves the accuracy and measurement capability of the FBG temperature sensor, the Wavelet Denoising Approach presented in the literature so far and the results of the proposed approach are compared. As a result, it is concluded that the Hilbert transform approach definitely follows better the true central Bragg wavelength values and shows smaller a relative error.