平面分层媒质中二维非均匀结构的电磁逆散射问题

本文在研究平面分层媒质中二维非均匀结构的电磁散射的基础上，研究埋藏在平面分层媒质中的二维非均匀结构的电磁逆散射问题。     

电磁逆散射非相关的照射法的普遍公式

由第一类Ｆｒｅｄｈｏｌｍ积分方程所描述的电磁逆散问题是非适定问题。通过一系列非相关照射，有可能获得射体的更多信息，从而改善解的非适定性。本文基于求解逆散射问题的非相关原理，推导了适于散射体在各种正交基下展开，求解电磁逆散射的普遍公式，并对所建立的参数反演公式解的唯一性进行了证明。     

电磁逆散射非相关照射法的普遍公式

由第一类Fredholm积分方程所描述的电磁逆散射问题是非适定问题。通过一系列非相关照射,有可能获得散射体的更多信息,从而改善解的非适定性。本文基于求解逆散射问题的非相关原理,推导了适于散射体在各种正交基下展开,求解电磁逆散射的普遍公式,并对所建立的参数反演公式解的唯一性进行了证明。     

2012年电磁散射与逆散射学术交流会征文通知

2012年电磁散射与逆散射学术交流会拟于2012年11月15-17日在四川省成都市召开。本次会议由中国电子学会天线分会电磁散射与逆散射学组主办,电子科技大学电子工程学院协办。这次会议将为国内电磁散射与逆散射领域的专家学者     

基于迭代多尺度深度网络的非线性逆散射成像方法

传统的迭代多尺度方法 (iterative multiscaling approach, IMA)在求解非线性电磁场逆散射问题时,可以自适应提高成像空间的分辨率,缓解逆问题的病态性,但容易陷入局部极小值且无法做到实时重构.文中提出了一种迭代多尺度深度网络,该网络结合传统IMA和深度网络的优势,将IMA展开成深度网络模型（命名为IMA-Net）.该方法迭代地执行一种感兴趣区域(regions of interest, ROI)提取算法,在不同尺度的ROI内构建目标函数,并将目标函数分解成若干个优化子问题,子问题的迭代更新过程映射到深度网络结构中,交替更新相关分量,求解出目标函数的最优值.实验结果验证了该方法的有效性和优越性,为目标实时重构提供了一个有效方案.     

电磁逆散射无相位检测的相位复原

利用入射场作为约束条件,用最优化泛函的方法由总场幅度信息重建散射场幅度和相位.分析了散射场自由度和噪声对重建结果的影响.讨论了成功实施相位复原和避免局部极小值的条件.首次用实验结果验证电磁逆散射无相位检测相位复原的可行性,证明了方法的有效性、高度稳定性和较强的抗噪声能力.该方法的计算成本低,实用性较强.研究结果是实施无相位检测电磁逆散射的基础.     

有限深度海域海面电磁散射的FDTD 方法研究

雷达能够探测、遥感有限深度海域的海面及附近目标,针对这一需求,采用改进的文氏海谱和Monte Carlo 方法模拟有限深度海域海面,选取Debye 模型计算海水介电常数,运用时域有限差分法研究了有限深度海域海面 电磁散射,仿真了散射系数的角分布曲线。仿真结果表明:散射系数随散射角振荡变化,在镜反射方向处产生散射增强 效应;海水深度虽然对散射系数影响较小,但是仍有规律可循,海水深度越大,散射系数越小;海面风速、入射角、风区范 围对散射系数影响较大,风速越大,散射系数越小;入射角越大,散射系数越小;风区范围越大,散射系数越小。     

深度学习下的散射成像:物理与数据联合建模优化（特邀）

为了利用被散射的光信号实现成像,越来越多的散射成像方法被提出。其中深度学习以其强大的数据表征和信息提取能力在散射成像领域发挥着重要的作用。相较于传统散射成像方法,基于深度学习的散射成像方法在成像速度、质量、信息维度等方面都有着巨大的优势。但是,模型训练、模型泛化等问题也制约着该方法的发展。因此,越来越多的研究将物理过程与基于数据驱动的方法进行联合建模,利用物理先验指导神经网络优化。相较于单纯的数据驱动方法而言,物理-数据联合建模的方法对数据量、神经网络参数量的依赖程度大大降低,在保证成像质量的前提下有效降低数据获取难度及对实验环境的要求。联合建模优化的方式实现了介质、目标类型等散射成像中关键节点的泛化。同时在训练过程方面,实现了从有监督到半监督再到无监督的训练优化过程迭代,不同模型和监督方式的提出大大提升了基于深度学习方法的训练效率,在降低对硬件和时间成本的同时,提升了基于深度学习的散射成像方法在非实验室场景应用的可能性。     

Dynamic differential evolution strategy and applications in electromagnetic inverse scattering problems

A novel evolution algorithm, the dynamic differential evolution strategy, is developed to solve optimization problems. It inherits the genetic operators from the differential evolution strategy. However, it differs itself from the differential evolution strategy by updating its population dynamically while the differential evolution strategy updates its population generation by generation. The dynamic updating of population leads to a larger virtual population size and quicker response to change of population status. Two trial functions have been minimized using the dynamic differential evolution strategy. Comparison with the differential evolution strategy has been carried out. It has been observed that the dynamic differential evolution strategy significantly outperforms the differential evolution strategy in efficiency, robustness, and memory requirement. It is then applied to solve a benchmark electromagnetic inverse scattering problem. The outstanding performance of the dynamic differential evolution strategy is further consolidated.     

Transform approach to electromagnetic scattering

In this paper, we present a comprehensive review of the Fourier transform technique as applied to the problem of high-frequency scattering and introduce the concepts of the spectral theory of diffraction (STD). In contrast to the more commonly employed ray-optical method for high-frequency scattering, viz. the geometrical theory of diffraction (GTD), the STD approach interprets the scattered field as the spectrum, or the Fourier transform of the induced current on the scatterer. Such an interpretation offers several important advantages: uniform nature of representation, capacity to improve and extend the ray-optical formulas in a systematic manner, and convenient accuracy tests for the results. Methods for combining integral equation methods with the Galerkin procedure and asymptotic techniques in the transform domain are described, and representative examples illustrating the application of the spectral approach are included.     

Reciprocity, discretization, and the numerical solution of directand inverse electromagnetic radiation and scattering problems

The author gives a formulation, based on Lorentz reciprocity, that unifies the finite element method (FEM) and the integral equation models. Wave propagation and scattering problems in electromagnetics have to be addressed with the aid of numerical techniques. Many of these methods can be envisaged as being discretized versions of appropriate weak formulations of the pertinent operator (differential or integral) equations. For the relevant problems as formulated in the time Laplace-transform domain it is shown that the Lorentz reciprocity theorem encompasses all known weak formulations, while its discretization leads to the discretized forms of the corresponding operator equations, in particular to their finite-element and integral-equation modeling schemes. Both direct (forward) and inverse problems are discussed     

Nonuniqueness in inverse source and scattering problems

The fields radiated by spherically symmetric time-harmonic sources are used to illustrate how little can be learned about a source from knowledge of the radiated field outside of the source volume. It is shown that even if it is known that the source is spherically symmetric, it is not possible to determine its radial structure. Moreover, even if the radial structure of the source is known apart from a constant amplitude and a finite radius, it is not possible to evaluate those two unknowns independently. These examples are applied to demonstrate explicitly that two methods that have been claimed to produce exact unique solutions to inverse-source and inverse-scattering problems do not yield the claimed results.     

Spectral and time domain approaches to some inverse scattering problems

Four methods of electromagnetic probing of an inhomogeneous plane stratified medium are presented. They differ by the nature of the incident wave used in the diagnostic. In the first method, the medium is considered as a structure propagating guided modes, and permittivity profiles are determined from the knowledge of one of the guided modes. In the next three, the medium is considered as a scatterer illuminated by a plane wave. Two of them are devoted to reconstruction of conductivity or permittivity profiles from time domain analysis of the reflected field. The last one allows one to obtain permittivity profiles from spectral domain analysis of the reflection coefficient. In spite of their apparent diversity, all these methods are issued from integral representations of the electromagnetic fields. These representations provides the simulated experimental data for the inverse problems. The behavior of these processes has been examined when parameters of practical interest are varied.     

A fast convergent method in electromagnetic inverse scattering

The authors present a fast convergent technique of the Born iterative method for nonlinear two-dimensional lossless object profile inversion using time-domain data. The magnitudes of the scattered fields at different frequencies are weighted according to the properties of the object. It is shown that the technique can speed up the convergence of the Born iterative method     

Piecewise smooth models for electromagnetic inverse problems

This paper presents a new method for constructing one-dimensional (1D) electrical conductivity models of the Earth from surface electromagnetic measurements. The construction of these models is a nonlinear inverse problem that can be approached by linearization techniques combined with iterative methods and Tikhonov's regularization. The standard application of these techniques usually leads to smooth models that represent a continuous variation of conductivity with depth. In this work, the authors describe how these methods can be modified to incorporate what is known in computer vision as the line process (LP) decoupling technique, which has the ability to include discontinuities in the models. This results in piecewise smooth models that are often more adequate for representing stratified media. They have implemented a relaxation technique to construct these types of models and present numerical experiments as well as an application to field data. These examples illustrate the performance of the combined LP and Tikhonov's regularization method     

Application of Haar-wavelet-based multiresolution time-domainschemes to electromagnetic scattering problems

The multiresolution time-domain (MRTD) algorithm is applied to the problem of general two-dimensional electromagnetic scattering. A Haar wavelet expansion is utilized. A parallel between Haar MRTD and the classic Yee finite-difference time-domain (FDTD) algorithm is discussed, and results of simulations on canonical targets are shown for comparison. We focus on the incident-field implementation, which, in our case, consists of a pulsed plane wave. Also, we consider scattering in a half-space environment, with application to subsurface sensing. The results illustrate the advantage of the Haar MRTD method as compared with the classic FDTD, which consists of reduced memory and execution time requirements, without sacrificing accuracy     

Digital realisation of forward and inverse models of electromagnetic scattering

A recent geometrical treatment of forward and inverse scattering problems, posed by electromagnetic wave illuminating a one-dimensional dielectric inhomogeneity, is interpreted within the framework of digital signal analysis. Quantitative inverse solutions are described, and are illustrated by an example in the context of real-time nondestructive evaluation.<>     

THE INVERSE ELECTROMAGNETIC SCATTERING OF INHOMOGENEOUS DIELECTRIC THIN TUBES

By using the method of Fourier series expansion, the inverse electromagnetic scattering problemof a 2- dimensional lossy dielectric body, of which the cross section is a narrow annulus, is studied. Theanalytic expressions of the electric parameters of a dielectric body are given in terms of the outsidesca,ttering field, and the simulative results are satisfactory.