Detection of Output of Fiber-Optic Bragg Grating Sensor Using Parallel Factor Analysis

In this paper, a method employing PARAFAC (Parallel Factor Analysis) is proposed to filter out noise from the signal of the FBG (Fiber Bragg Grating) sensor based on the AWG (Arrayed Waveguide Grating). The experimental results show that the performance of the proposed method is similar to that of the conventional signal-averaging method, which identifies the signal in the case of multiple measurements. PARAFAC is one of several decomposition methods for three-way data and the decomposition is unique without additional orthogonality or independence constraints. When Bragg wavelength of the FBG varies upon receiving an ultrasonic wave, the outputs of two adjacent AWG ports fluctuate in anti-phase. We use wavelet transform to pre-process the AWG ports data series, then the coefficients of the wavelet transform are decomposed by PARAFAC to trilinear components, which represent frequency, temporal, and phase profiles. The factor for which the ports fluctuate in anti-phase is chosen as the target signal and restored to three-way data. The performance of the proposing method is examined in experiments and its validity is shown.