基于散射中心模型的雷达目标联合跟踪与识别技术研究

Joint processing of target tracking and recognition is an important trend in military electronic equipment, and it is also a hot topic in the research field of radar detecting. Currently, research work is mainly carried out based on the assumption of point target model and the simple classification feature (the quantity of acceleration) represented by the maneuvering mode, and the applicability of developed approaches is very limited. Meanwhile, because only the kinematics measurements are used for joint estimation, the improvement of target recognition ability is also very difficult. The rigid-body model provides an effective way for generating the complicated feature of target recognition; however, it is only suitable for optical sensor application. In this project, aimed at the great requirements of radar detecting for the sea/ground target, the modeling technology of scattering characteristic for complex configuration target will be deeply investigated by starting from the geometrical (CAD) model, and the 3D scattering center model will be reconstructed from the predicted RCS data under different target pose and electromagnetic wave frequencies, thus to realize the dynamic modeling of radar echoes and the on-line generation of recognition features. The estimation of target state and type are then jointly obtained by incorporating the kinematics measurements, electromagnetic measurements and nonlinear inference approaches. The performance enhancement techniques will also be full considered in the work research, and what achieved in the project will provide important supports of theory, model and algorithm for the development of integrated detection technology.

目标跟踪与识别联合处理是军用电子系统技术发展的重要趋势，也是雷达探测识别领域的研究热点。现有的研究主要是针对质点模型和机动模式（加速度的大小）这一简单识别特征进行的，适用范围非常窄；同时，由于仅利用了运动学量测信息，也不利于目标识别能力的提升。 刚体模型为复杂识别特征的利用提供了可能，但这种模型只适用于光学探测条件。针对雷达对地海面运动目标探测识别的重大需求，本项目拟从典型目标的几何模型出发，研究复杂形体目标的电磁散射特性建模技术，通过不同姿态下的宽带电磁散射特性预估数据构建出目标的三维散射中心模型，从而实现复杂形体目标的雷达动态回波建模与识别特征的在线生成。在此基础上，结合目标运动学特性、电磁散射特征量测及非线性系统推理技术，获得目标运动状态与类属的联合估计，并通过多量测信息融合手段提高目标跟踪与识别的联合处理性能，为集成化/一体化的雷达目标探测识别技术发展提供理论、模型和算法支撑。

本项目主要针对雷达对地海面运动目标探测识别的重大需求，重点研究复杂形体目标的三维散射中心建模技术、运动目标的识别特征预测技术、基于模型的目标跟踪与识别联合处理技术，以及基于信息融合的目标跟踪与识别联合增强技术等。项目组通过4年的深入研究，突破了雷达目标三维散射中心模型重构技术，实现了复杂结构目标的动态回波模拟与特征预测，结合目标运动学特性、雷达散射特征量测及非线性系统推理技术，获得了目标运动状态与类属的联合估计，并通过多视观测距离像、极化距离像等信息的利用，增强了目标的识别能力，为集成化/一体化的雷达目标探测识别技术发展提供了理论、模型和算法支撑。.此外，通过本项目的实施，建立了3类典型海面舰船目标的几何模型和散射中心模型，仿真获取了4类典型地面车辆目标的全极化、方位全姿态高分辨距离像数据，构建了支撑联合目标跟踪与识别技术研究和性能验证的基础数据集。在国内外重要期刊发表论文11篇（SCI检索7篇，Ei检索10篇），在国际会议发表论文5篇（其中Ei检索4篇，ISTP检索4篇），申请国家发明专利2项。通过本项目的支撑，培养博士研究生4名、硕士研究生3名（均顺利毕业），成长领域内青年技术骨干1名。