基于差分演化算法的啁啾光纤光栅参数重构方法

Parameter reconstruction of chirped fiber grating using differential evolution algorithm

本文利用差分演化算法重构了啁啾光纤光栅的结构参数。根据光纤光栅的目标反射谱特点,利用该算法搜索最优的一组光栅参数,根据这组参数计算得到的理论反射谱相对于目标反射谱的偏差最小,由此得到优化的啁啾光纤光栅结构参数。数值实例表明重构参数与目标参数非常接近,相对误差非常小。与遗传算法的离线性能比较表明,差分演化算法需要的迭代次数少,收敛速度快,在有大量矩阵运算的啁啾光纤光栅参数重构这类问题中具有更高的计算效率和更强的应用性。

The structural parameters of chirped fiber gratings are reconstructed based on the target reflection spectra using differential evolution optimization algorithm. The calculated reflection spectra associated with reconstructed parameters show good agreement with the target reflection spectra. Comparing with genetic algorithm, the differential evolution algorithm is more effective and practical for parameter reconstruction of fiber gratings.