多基地雷达空间目标惯量特征分析与提取方法研究

The ground-based radar is one of the most important measures for space target acquisition and recognition. In this National Natural Science Foundation of China (NSFC) grant, we shall investigate the inertial characteristic estimation of free rigid targets in space by using multistatic narrowband radar. For wide applicability of the research results and without loss of generalization of space objects, the targets with the mass asymmetry are taken into account in the study. First, based on the attitude kinematics in space, some smart simplification and transformation will be applied to the inertial model of the targets. It could be clearer to specify the relationship between targets’ inertial characteristics, micro-motion states and multistatic radar micro-Doppler features, which plays a key role in the parameter estimation of inertial characteristics afterwards. Besides, the Doppler measurement in multistatic radar will be considered, and together with the improved inertial model, we will analyze the time variation of multistatic radar micro-Doppler from asymmetric targets. Second, the spatial diversity and multistatic radar micro-Doppler will be exploited for the parameter estimation of inertial characteristics of asymmetric targets in space. After time-frequency analysis and extraction, we could have a series of micro-Doppler time-frequency data sequences related to multi-channel signals and multi-scatterer on target. Then, the self-similarity matrix and the global nearest neighbor criterion will be used to achieve the measurement-to-measurement micro-Doppler association. On this basis as well as the physical relationship mentioned above, the inertial characteristic parameters of asymmetric target in space could be estimated from multistatic radar echoes. Third, the performance assessment will be conducted by means of theoretical analysis, simulations and software-based electromagnetic scattering data. In particular, a new multistatic radar ambiguity function will be derived and discussed with the consideration of radar system configuration in three-dimensional space and high-order approximation of micro-motion. The oncoming research achievements in our grant proposal will contribute to the non-cooperative space target recognition in radar, technical upgrade for active-duty narrowband radars and multistatic radar system configuration, and be applied in space exploration, support to manned space missions, homeland defense, and so forth.

地基雷达探测是空间非合作目标测量与识别的重要途径之一。本项目将以多基地窄带信号体制雷达为手段，针对空间刚体目标质量非对称分布的一般化情况，研究利用多基地雷达空间分集特性，通过自相似矩阵和全局最近邻准则对多路回波微多普勒时频分量进行关联处理，实现目标惯量参数估计的方法。根据空间动力学原理和多基地雷达多普勒测量特点，分析目标惯量参数与微动状态及微多普勒特征之间的明确量化关系，在理论、算法和实验三个层面上开展对惯性模型合理简化、目标惯量参数估计、性能评估方法等方面的研究工作。项目研究成果将为雷达目标特征提取与识别提供新的思路和方法，并能够为窄带信号体制雷达技术升级、多基地雷达系统合理构建提供关键的理论支持，服务于我国在空间探测、载人航天保障、导弹防御等领域的技术发展需求。

空间目标除轨道平动外，由于初始横向干扰的影响将导致其自身产生姿态运动，即微动。通常情况下，由这些微小运动引起的雷达微多普勒频率调制现象不仅反映了目标的精细特征，而且还能够体现目标内在的固有属性信息。本项目主要以多基地窄带信号体制雷达为手段，针对空间刚体目标质量非对称分布的一般化情况，研究了目标微动及其对雷达回波调制的影响；提出了基于全局最近邻和混合微多普勒时频序列的空间目标惯量特征估计方法，以及基于多基地雷达距离历程的目标强散射中心三维分布估计方法；在此基础上，分析了雷达布站对目标微动及惯性参数估计的影响。.首先，推导和分析了质量非对称分布空间刚体目标微动以及相应的雷达回波微多普勒瞬时频率特征，表明在一般情况下，目标微动状态在本质上与自身转动惯量和初始起旋角速度有关。因此，微多普勒频率的时变特征能够为利用雷达回波进行目标微动及惯量特性的反演提供有效的手段。.其次，针对多微多普勒分量的数据关联问题，研究了基于全局最近邻和混合微多普勒时频序列的空间目标惯量特征估计方法。利用时频分量的全局概率和卡尔曼滤波来进行其更新与预测。在此基础上，通过加权叠加形成一维混合微多普勒时频序列，并采用基于傅里叶变换的渐进最大似然估计方法进行频谱估计，实现对空间目标惯量特征的有效提取。.再次，针对质量非对称分布空间刚体目标微动形式复杂的问题，讨论利用多基地雷达观测，实现对目标散射中心空间三维分布的有效估计。在构建多散射中心距离历程的混合测量矩阵基础上，利用奇异值分解，并设计响应的约束条件，实现对质量非对称分布空间刚体目标强散射中心三维分布的确定性估计。.最后，针对空间目标回波微多普勒瞬时频率与雷达观测角度之间的关系，研究了雷达布站对目标微动及惯性参数估计的影响。利用微多普勒频率带宽描述了雷达观测角变化与瞬时频率动态范围的关系，分析了雷达观测视角对混合微多普勒时频序列中各频谱分量盲提取和匹配处理的影响。