极化特性是弹道目标识别的重要特征之一,针对复杂突防环境下弹头快速识别需求,提出了一种基于分时极化重构和特征压缩的弹头识别方法。利用微动目标的极化散射矩阵暗室测量数据,分析了动态目标的极化散射特性。结合弹道轨道模型,仿真了弹道目标的动态极化散射矩阵雷达回波信号。基于运动参数和极化互易性对极化矩阵进行相位补偿重构;然后利用重构的极化矩阵提取极化不变量特征和庞加莱球点的变化等特征,并采用多维特征投影变换实现特征压缩,去除无用冗余信息;最后通过软判支持矢量向量机进行序贯识别,从而实现对弹头、碎片和发动机等弹道目标的快速稳健识别。

Polarization characteristic is one of the important characteristics of ballistic target recognition. Aiming at the requirement of rapid warhead recognition in complex penetration environment, a warhead recognition method using time-division polarization reconstruction and characteristic compression is proposed in this paper. The polarization scattering characteristics of the dynamic target are analyzed, using the polarization scattering matrix anechoic chamber measurement data of the micro-motion target. The dynamic polarization scattering matrix radar echo signals of the ballistic target are simulated and analyzed. Based on the motion parameters and polarization reciprocity, the polarization matrix is reconstructed with phase compensation. Then, the reconstructed polarization matrix is used to extract the characteristics of polarization invariant features and changes of Poincaré ball points, and the multi-dimensional feature projection transformation is used to realize feature compression and remove useless redundant information.Finally, sequential recognition is carried out through soft decision support vector machine, so as to realize fast and robust recognition of ballistic targets such as warheads, fragments and engines.

TN957