求解本征值问题的FDTD快速算法

Fast FDTD Algorithms for Solution of Eigenvalues

本文分别引入数字信号处理中的Ｍａｒｐｌｅ算法，神经网络的学习算法，与ＦＤＴＤ法结合，形成两种新的快速算法，与现有的ＭＵＳＩＣ算法比较，着重讨论了三种算法的优缺点．并用三种算法求解复杂微波结构的本征值问题，计算结果与未利用数字信号处理的结果相比，快速算法很大程度地提高了ＦＤＴＤ法的效率和分析精度．

In this paper,Marple algorithm and learning algorithm are introduced to develop two novel fast FDTD analysis algorithms.The difference among the three algorithms including the extant MUSIC algorithm is presented.With the three algorithms,a complicated microwave structure is analyzed.Compared with the result without signal processing,a great improvement of the efficiency and precision of FDTD method is achieved.