THE ELEVENTH IEEE SENSOR ARRAY AND MULTICHANNEL SIGNAL PROCESSING WORKSHOP (IEEE SAM2020)

6 / 2019-11-18 08:41:00

Denoising of the Fiber Bragg Grating Deformation Spectrum Signal Using Variational Mode Decomposition Combined with Wavelet Thresholding

fiber Bragg grating; variational mode decomposition; discrete wavelet transform; signal processing

IEEE SAM 2020: Full Regular Papers > Detection and estimation

Damage detection using a Fiber Bragg gratings (FBG) sensor is a critical process for an assessment of any inspection technology classified as structural health monitoring (SHM). FBG signals containing noise in experiments are developed to detect flaws. In this paper, we propose a novel signal denoising method that combines variational mode decomposition (VMD) and changed thresholding wavelets for mixed signals denoising. VMD is a recently introduced adaptive signal decomposition algorithm. Compared with traditional empirical mode decomposition (EMD), and it is well founded theoretically and more robust to noise samples. First, input signals were broken down into a given number of K band-limited intrinsic mode functions (BLIMFs) by VMD. For the purpose of avoiding the impact of basic modes on VMD denoising, the mixed signals, which were obtained by adding different signal/noise ratio (SNR) noises to the experimental signals, were designed to select the best decomposition number K and data-fidelity constraint parameter α. After that, the realistic experimental signals were processed using four denoising algorithms to evaluate denoising performance. The results show that, upon adding additional noisy signals, the proposed algorithm delivers excellent performance over the EMD-based denoising method and discrete wavelet transform filtering.