FDTD 结合支持向量机反演各向异性材料电磁参数

本文首先介绍了太赫兹波导和3D打印技术的发展现状。3D打印作为一项新兴的技术,以数字模型文件为基础,运用粉末状金属或塑料等可粘合材料通过逐层打印的方法构造实体,打破了传统THz波导技术的局限性。本文介绍的3D打印THz波导利用聚合树脂作为打印材料,打印完成的THz波导在其传输通路上镀500nm的金,金的厚度足以支持THz传播。利用这种方法可以打印出直波导、三维弯曲面、三维Y劈和U型波导等多种结构。3D打印THz波导除传输损耗略高外,其传输模式及其特性与传统的金属波导基本一致,这种额外的传输损耗归咎于商业3D打印机的精度。

Inversion of Electromagnetism Parameters of Anisotropic Material Using Support Vector Machine and FDTD Method

In this paper, the machine learning method-support vector machine(SVM) is used to invert the electro-magnetism parameters of anisotropic materials . To this aim, the radar cross section(RCS) of anisotropic medium sphere are numerical simulation by time-domain finite different(FDTD) method, and provide to SVM as input training samples. After proper training, the relative dielectric constant and electric conductivity of the anisotropic medium sphere are reconstructed in real time. At the same time, we also give the results by using BP neural network method , shows the validity and accuracy of the presented method.