节点文献

目标与粗糙面复合电磁散射特性的快速数值算法建模

Simulation of Electromagnetic Scattering Characteristics from Composite Model of Target and Rough Surface Using Fast Numerical Method

【作者】 邓方顺

【导师】 朱国强;

【作者基本信息】 武汉大学 , 无线电物理, 2010, 博士

【摘要】 海面(或地面)环境中目标电磁散射特性,对实际战场环境下军用目标的雷达探测、雷达特性形成以及雷达目标识别等技术具有基础支撑作用;因此,粗糙面环境下目标雷达特性建模与算法研究自然成为既具有重要理论意义,又具有重大实际应用价值的现代计算电磁学前沿应用课题之一。本文分别针对二维和三维目标与粗糙面复合电磁散射问题,在矩量法(MOM)的框架下开展建模和快速数值求解方法研究。求解二维标量波散时,引入多层UV (MLUV)的矩阵分解技术和稀疏矩阵规范网格方法(SMCG)方法加速求解过程,包括矩阵元素快速填充及矩阵方程的快速迭代求解。在三维目标和粗糙面复合电磁散射特性建模时,采用基于表面电流的矢量基函数(RWG)离散表面积分方程,针对此时震荡积分核使传统的UV分解技术失效的问题,提出了基于电磁相互作用的排序粗取样UV分解算法,其中等弧长粗取样算法的自适应特性能满足各种复杂电磁相互作用精确计算的要求,有效地解决了因震荡积分核导致UV分解失效的问题。将原来的标量UV算法发展为三维矢量UV算法,形成了粗糙面环境下目标电磁散射建模的高效、高精度三维矢量多层UV (3DMLUV)建模方法。文中详细分析了目标与粗糙面之间复杂电磁相互作用的机理与对应的物理过程,并通过仿真实例验证了本文建模方法的准确性、有效性。本文主要内容包括:1)研究了二维目标与一维随机粗糙面复合电磁散射特性的MLUV-SMCG混合算法建模。并针对矩阵方程迭代收敛慢的问题,提出了快速代入迭代(FSIA)过程加速求解,形成了快速的MLUV-SMCG-FSIA混合方法。数值模拟了一维导电P-M海谱粗糙面上方导电圆柱和方柱目标的双站雷达散射截面(Bi-RCS),比较和分析了不同极化入射波下的粗糙面上方不同电尺寸目标的复合电磁散射特性。2)研究了PEC情况下三维目标与二维随机粗糙面复合电磁散射特性的算法建模,针对三维矢量波散射,发展了原有的标量UV方法,提出了基于电磁相互作用的高效、精确的3DMLUV建模方法。数值模拟了二维随机粗糙面上方导电立方体、球体以及粗糙面上三维简化船只目标的Bi-RCS;通过与导电平板上相同目标复合散射特性的比较,对电磁相互作用特性进行具体分析;此外还讨论了粗糙面的粗糙度对复合散射的影响。3)研究了二维介质粗糙面上任意三维介质/PEC目标复合电磁散射特性的算法建模,针对介质问题,结合快速互耦迭代方法(FIA),提出了3DMLUV-FIA算法建模。数值模拟了单独介质体、单独介质粗糙面和体/面组合散射的双站雷达散射截面,通过与MOM和其它快速算法的比较,表明在介质散射问题求解中,3DMLUV-FIA算法具有稳定、准确和高效的特点。

【Abstract】 The study of composite electromagnetic scattering from target and rough surface combined model has wide applications in radar target identification and microwave remote sensing. Therefore, simulation of scattering characteristics from the target under the rough surface environment is of great value in both theory study and practical applications. The scattering from the large size target and rough surface composite model for both two-dimensional and three-dimensional problems is investigated in this paper. For two-dimensional problem, based on the MOM,the multilevel UV (MLUV) matrix decomposition technique is combined with sparse matrix canonical grid (SMCG) method to implement the fast computation of impedance matrix elements and the iterative solution of matrix equation. While for three-dimensional vector wave scattering problems, an EM-interaction-based sampling algorithm is developed to overcome the difficulties brought by the oscillation of the interaction matrix elements which invalidate the conventional UV decomposition technique. The numerical simulation includes the composite radar cross section (RCS) evaluation and the radar wide-band imaging of the target and rough surface model; the complex electromagnetic mutual interactions between the target and the rough surface are analyzed in details and the hybrid method proposed in this paper is verified through the numerical examples.The main content of this paper includes the following parts:1. The hybrid MLUV-SMCG-FSIA method is proposed for the composite electromagnetic scattering from the 2-D target and rough surface combined model. The bistatic radar cross section (Bi-RCS) of the cylinder and the square cylinder above the P-M rough sea surface is numerically simulated and the composite scattering from different size targets above rough surface under different incident wave polarization is compared and analyzed.2. The 3-D multilevel UV method (3DMLUV) is proposed for the vector wave scattering from PEC target and rough surface composite model. The Bi-RCS of a cube and a sphere above the random rough surface and a simple 3-D ship on the rough surface is numerically simulated and compared with target and the planar surface model case. The influence from the roughness of the underlying surface to the scattering is discussed.3. The 3DMLUV-FIA method is proposed for the vector wave scattering from the 3-D PEC or dielectric target and dielectric rough surface composite model. The Bi-RCS of single dielectric target, single dielectric rough surface and volume/area combined scattering is numerically simulated. The characteristics of 3DMLUV-FIA method in stability, accuracy and efficiency are demonstrated by comparing with the full method of moment (MOM) and other fast algorithms.

  • 【网络出版投稿人】 武汉大学
  • 【网络出版年期】2010年 10期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络