Efficient two-stage strain/temperature measurement method for BOTDA system based on Bayesian uncertainty quantification

Brillouin optical time-domain analysis (BOTDA) is an efficient distributed structural strain/temperature mea-surement technology by identifying the Brillouin frequency (BF) of the fiber. The original BF identification method involves gradually tuning the scanning frequency of the probe laser light, which is time-consuming. To improve the sensing efficiency of BOTDA, a novel two-stage strain/temperature measurement method is pro-posed based on Bayesian uncertainty quantification. The method uses the information of a small amount of wide range frequency scanning data in the 1st stage to guide a concentrated scanning of the frequency points in the 2nd stage by employing transitional Markov chain Monte Carlo (TMCMC) algorithm. The second-stage scanning strategy based on Bayesian uncertainty quantification ensures the high confidence for BF identification, so that the number of required scanning points can be significantly reduced without sacrificing the measurement ac-curacy. Finally, the efficacy of the proposed method is verified using numerical and experimental data.