基于多层平面波分解的快速近远场转换新方法研究

The Near-Field Far-Field Transformations (NFFFT) of target electromagnetic scattering is important in the Radar Cross Section (RCS) measurement in Near-Field. The traditional NFFFT methods include the plane-wave synthesis based methods and the scattering center distribution based methods. The plane-wave synthesis based methods are hard to solve the electrically large targets’ scattering problem because of the enormous measurement efforts and high computational cost. The scattering center distribution based methods work with broadband monostatic observations on sphere or plane or cylindrical surface, the need with the fast measurement scene still inconformity. The scattering center distribution based methods cannot handle the strong coupling effect in target because the high-order scattering component is ignored. The extra antenna pattern compensation is needed in both the two traditional NFFFT methods. In the project we present a fast NFFFT method by multilevel plane-wave decomposition, which is derived from a spherical expansion of the near-field according to addition theorem, and rather accelerate the plane wave modes expanding using multilevel fast multipole method (MLFMM). The fast method can apply to the full model with full bistiatic measurements, and can also apply to the linear model with monostatic measurements. There are many advantages of the fast method: the MLFMM acceleration reduces the computational cast; the great flexibility with respect to measurement locations and probes, because such effects are fully compensated in the iterative inversion algorithm; the sampling number is the same as the multipole order, it’s a nonredundant sampling scheme on arbitrary surfaces; the self-contained theory system make it easy to control and analysis the error. In the project, the theoretical derivation will do firstly. Second, we will numerically implement the MLFMM-based NFFFT method. At last, the experiments will be performed in the chamber.

目标电磁散射特性的近远场转换是通过近场测试获取目标雷达散射截面的关键技术。传统的近远场转换有基于平面波综合和基于目标散射中心两种方法：前者由于巨大的近场双站采样量和计算量，难以扩展至电大尺寸目标的散射问题；后者需要规则面上的宽频带单站采样，但不能解决强耦合目标问题。本项目提出的基于多层平面波分解的快速近远场转换方法，根据加法定理用多层平面波展开目标散射近场，用多层快速多极子算法加速球面波的平面波谱展开过程。这种方法既可用于基于三天线链式关系的双站采样完备散射模型，又可用于基于目标散射中心的单站采样线性散射模型，具有以下优势：使用多层快速多极子算法加速近远场转换过程，降低算法复杂度；测试天线和采样位置引起的影响完全在迭代求逆算法中得到补偿，极具灵活性；采样量与多极子展开阶数一致，达到非余量采样；理论体系完整，误差可控。本项目研究从理论推导入手，开展快速算法数值实现，最后在微波暗室中实测验证。

低电磁散射技术的深入应用对目标雷达散射截面（RCS）测试提出了现场、快速、便捷的要求。近场测试及近远场转换技术是在目标电磁散射近场获取其RCS的手段，因其对场地要求低受到越来越多的关注。本项目研究内容包括：基于平面波分解技术得出一组散射问题的完备的双站近远场转换公式；基于平面波分解及辐射-散射模型得到单站近远场转换方法；实现基于多层快速多级子技术的近远场转换算法；用近场测试模拟仿真数据和暗室测试数据验证近远场转换方法的正确性。主要研究结果如下：. （1）推导了完备的散射问题近远场转换公式，并在基于Born近似的辐射-散射模型基础上推导出了单站近远场转换公式，通过单站近远场转换和双站近远场转换之间的近似关系，论证了近远场转换精度与近场采样量之间的关系：近场采样量越大，近远场转换精度越高。双站采样是完备的，但测量量巨大难以工程实现；单站采样是基于线性近似的，理论上不适用于多次散射，但实际数据显示单站近远场转换的精度已足以满足工程应用需求。. （2）利用MLFMM算法实现了多层分组结构的快速近远场转换方法，并通过分析各层平面波谱的方相性进行了算法改进。算法复杂度达到级别。. （3）对于具有工程应用价值的单站近远场转换方法，通过简单目标的近场仿真数据和暗室测试数据验证了其精度具有和其他基于近场成像的近远场转换方法相似的转换精度，但是介于本项目方法可以在迭代过程中消除天线方向图影响，所以精度稍高。另外本项目方法具有任意点采样、单频、全极化的特点，从而测试效率更高。. 通过本项目的研究，得到在目标散射近场进行单站、单频、全极化、任意天线（非全照射）测量就可以转换得到远场RCS的结论，相对误差可控制在-40dB以下。. 本项目研究的科学意义是，证明有限边界源散射电磁场可以用任意表面上有限采样点精确表达，采样点的数目与平面波展开阶数一致。在项目研究结果指导下，可以构建一种不需要转台等大型设施的轻量型目标近场测试场，实现目标RCS现场测试。