轴对称二维介质体的非线性优化反演方法

Nonlinear optimize inversion for two-dimensional axix-symmetric medium objective

采用一种非线性优化反演方法———遗传算法结合FDTD方法求解轴对称二维介质柱电磁逆散射问题。定义目标函数为实测场与计算场之间的归一化误差,使用遗传算法最小化这个目标函数即可得到目标区域的介电参数分布,从而实现逆散射问题的求解。通过对参数基本设置的分析,总结出参数设置的经验。最后通过计算机模拟,给出几个反演算例,反演结果与真实目标吻合很好。

It proposes an effective method to reconstruct two-dimensional medium. The reconstruction is obtained from the interative inversion method that minimizes the cost function by using Genetic Algorithm, where the cost function is defined as the unitary error between the measured fields and the calculated fields from the assumed profile of the object. The scattered fields are calculated by the Finite Difference Time Domain （FDTD） algorithm. It summarizes the experience of how to intercalate parameters after analyzes parameters＇ setting. Numerical result shows that these methods are available of reconstructing a two-dimensional medium.