Abstract:In this paper, we present a simple and fast spectra inversion method to reconstruct the temperature distribution along single fiber Bragg grating (FBG) temperature sensor. This is a fully distributed sensing method based on the simulated annealing evolutionary (SAE) algorithm. Several modifications are made to improve the algorithm efficiency, including choosing the most superior chromosome, setting up the boundary of every gene according to the density of resonance peaks of the reflection spectrum, and dynamically modifying the boundary with the algorithm running. Numerical simulation results show that both the convergence rate and the fluctuation are significantly improved. A high spatial temperature resolution of 0.25 mm has been achieved at the time cost of 86 s.

Modified Simulated Annealing Evolutionary Algorithm for Fully Distributed Fiber Bragg Grating Temperature Sensing