Research of peak-detection algorithm in a fiber grating demodulation system

In order to improve the precision of peak-detection of a fiber Bragg grating demodulation system,it was studied based on the premature convergence characteristics of the genetic algorithm.It was expounded how the genetic algorithm can improve the accuracy of fiber Bragg grating demodulation.Taking a fiber Bragg grating demodulation system based on fiber Fabry-Perot(F-P)wavelength filter as example,the feasibility and reliability of the algorithm were testified.The peak reflected spectrum of the fiber Fabry-Perot filter fiber Bragg grating demodulation system was detected with MATLAB and its central wavelength was obtained.The effect of initial population number,genetic algebra,mutation rate on the peak detection of the algorithm was analyzed and a set of optimal parameters were obtained.The precision was equal or less than 3pm,standard deviation was less than 0.5pm for 50 repetition computations.The results show that the algorithm is stable,effective,and practical,can significantly reduce the noise of a fiber grating sensor system,improve the precision of demodulation.