基于多子群竞争进化规划算法的光纤光栅设计

从光栅传输矩阵分析法出发,提出了一种基于多子群竞争进化规划算法的光纤Bragg光栅(FBG)设计方法,进而设计了用于滤波的均匀FBG和用于色散补偿的啁啾FBG(CFBG)的结构参数.结果表明,本文设计方法可直接确定FBG的结构参数,在不同的设计精度要求下所设计的FBG分析曲线与目标曲线吻合很好,所采用算法具有比传统的进化规划算法快几10倍的收敛速度和更好的稳定性.

Fiber Bragg Grating Designing Based on Multi-subgroup Competition Evolutionary Programming Algorithm

This paper presented a simple,accurate,and fast method to synthesize the physical parameters of fiber Bragg grating numerically from FBG reflection spectrum. The method is based on the transfer matrix method and a Multi-subgroup competition genetic algorithm. The novel method was applied to design the structure parameters of uniform FBG for filtering,and chirped FBG for dispersion compensation. The analysis curve of the gratings shows a good agreement with the expected curve,and the algorithm is better than the calssical evolutionary programming algorithm in the aspects of global optimization and convergence.