双/多基地雷达空间微动目标联合特征提取技术研究

The precise physical feature extraction of those space targets with micro-motion, such as ballistic missiles, is of great significance for automation target recognition in the defense radar. The monostatic radar meets the problems of the self-occlusion and coupling of micro-motional and geometric parameters in feature extraction. Because bi-/multi-static radars can obtain the redundant information, more comprehensive target information can be extracted by information fusion, which will be more useful for target identification. This project focuses on the combined technique for feature extraction of space targets in bi-/multi-static radars. Firstly, the space-frequency characteristics of bistatic scattering of space targets are studied in-depth, combined with the orbit and micro motions of space targets, the modulation mechanism of bistatic radar echo of the spatial micro-motional target is revealed; secondly, in view of the influence of the target structure, visual conditions, geometrical configuration, the synchronization error and other practical factors on bistatic micro-Doppler, the robust estimation method of bistatic micro-Doppler based on energy scattering is proposed; finally, based on the association of the multi-aspect micro-motion and scattering characteristics, the combined feature extraction method of spatial targets is studied, which provides the meaningful feature corresponding to the physical attributes. This project is of great significance for space target recognition with bi-/multi-static radars.

弹道导弹等空间微动目标精细的物理特征提取对反导雷达的自动目标识别具有重要意义。单基地雷达在空间微动目标特征提取中面临自身遮挡以及微动参数和几何参数耦合的难题，而双/多基地雷达能获得冗余信息，通过联合特征提取能获得更丰富、更稳定的特征量，更有利于目标识别。本项目着重研究双/多基地雷达空间微动目标联合特征提取方法。首先，深入研究空间目标双基地散射的空频特性，结合空间目标质心运动和微运动，深入揭示空间微运目标双/多基地雷达回波调制机理；其次，在分析目标结构、可视条件、几何配置、同步误差等因素对双基地微多普勒的影响基础上，研究了基于能量扩散特性的双基地微多普勒估计方法；最后，研究了基于多基地多视角微动特性和散射特性关联的联合特征提取方法，提取物理意义明确的特征量。本项目对于双/多基地雷达空间目标识别具有重要意义。

以双/多基地雷达在空间目标监视中的应用需求为背景，深入研究了双/多基地雷达空间微动目标特征提取技术，重点开展了空间目标双基地散射特性分析与动态回波建模，双基地微多普勒检测与估计，双/多基地联合特征提取三方面的工作。首先，利用理论推导、电磁计算和暗室测量等手段获取目标双基地散射特性数据，并提出了空间目标全极化双基地动态回波仿真方法，充分体现了双基地回波调制机理。在此基础之上，将微多普勒曲线视为目标运动轨迹，建立了目标运动模型和观测模型，提出了基于粒子滤波检测前跟踪算法的微多普勒曲线检测方法，并用电磁散射数据动态仿真回波进行了验证。能适应目标自身遮挡、信噪比低等恶劣条件。考虑空间微动目标实际的双基地散射特性，同时利用目标微动特性和结构特性进行参数关联，提出了一系列基于多视角时频分布、时间-距离分布、双基地ISAR序列的空间微动目标联合特征提取方法，能同时获得目标物理尺寸和微动特征，与传统的单基地雷达空间微动目标特征提取方法相比，所得特征量更丰富、更稳定。