ABSTRACT This paper is to reconstruct the periodic inhomogeneous dielectric distribution of the scatterer buried in the rough surface. In order to explore the dielectric coefficient distribution of the unknown dielectric object under the rough surface, we emit electromagnetic wave to the object and measure the scattered electromagnetic wave above the rough surface. Base on Green's identity and the induced current concepts, the nonlinear integral equation can be derived and solved by the method of moments. Next, the inverse scattering problem is converted into an optimization problem. We use Self-Adaptive Dynamic Differential Evolution (SADDE) and Asynchronous Particle Swarm Optimization (APSO) to find the extreme value of the problem. With the regularization technique, the reconstruction is good. When the noise is less than 1%, the dielectric constant can also be achieved successfully. Numerical results show that SADDE can reduce the error for the permittivities of the object better than the APSO.

中文翻译：

---

周期性非均匀介电散射体的逆散射

摘要 本文旨在重建埋藏在粗糙表面的散射体的周期性非均匀介电分布。为了探究粗糙表面下未知介电物体的介电系数分布，我们向物体发射电磁波并测量粗糙表面上方的散射电磁波。基于格林恒等式和感应电流概念，可以用矩量法推导出非线性积分方程并求解。接下来，逆散射问题转化为优化问题。我们使用自自适应动态差分进化 (SADDE) 和异步粒子群优化 (APSO) 来找到问题的极值。使用正则化技术，重建效果很好。当噪声小于 1% 时，介电常数也可以成功实现。数值结果表明，SADDE 比 APSO 能更好地减少物体介电常数的误差。