Denoising of fiber grating sensor signals based on threshold retention orthogonal matching pursuit

Abstract

Aiming at the noise signal caused by external environment and sensor reuse in the demodulation system of large capacity fiber grating sensor network, a Threshold Retained Orthogonal Matching Pursuit (TROMP) reconstruction algorithm is proposed based on compressed sensing theory. In this algorithm, the useful signal can be effectively recovered from the noise signal by improving the threshold setting, atom selection and iteration stop condition. Experiments show that: TROMP algorithm can achieve ideal denoising and reconstruction effect under less observation value, the SNR of the system is increased by 10 dB, the root mean square error is less than 0.0071, and the number of cross-relations can reach 0.9999. Compared with similar algorithms, the running time is shortened by more than half, and the comprehensive performance of noise signal processing for FBG sensor system is better than other reconstruction algorithms.