空间群目标下多假设跟踪方法研究

发布时间：2019-01-20 10:48

【摘要】：弹道导弹在弹道中段通常有诱饵、干扰机及发射碎片等伴飞,形成一个庞大的威胁目标群,群内目标间隔不大,并保持大致相同的空间运动轨迹。地基雷达作为对中段目标的主要监测手段,在采用传统的多目标跟踪方法对密集空间群目标进行跟踪时,由于资源有限,造成个体目标的跟踪信息误差极大甚至缺失。因此,研究空间群目标中的单个目标航迹跟踪,对提高弹道导弹防御雷达的多目标跟踪能力,以及分析弹道导弹突防效能具有十分重要的意义。本文首先分析了弹道中段空间群目标的形成原因,并简述了这种特殊的密集群目标运动特点。其后,本文对常用的群目标跟踪理论进行了介绍和分析。对涉及目标跟踪中的数据关联算法和跟踪滤波算法,本文进行了简要介绍和分析,为后面空间群目标下的的多假设跟踪算法研究奠定了理论基础。其后,本文提出了基于群内聚类的群中心跟踪方法(IGC-SGT),显著提高了群整体跟踪的稳定性。针对群内目标的波动和误差严重影响群中心稳定性的问题,IGC-SGT算法改进了群内聚类方法,充分利用群内聚类的跟踪,获取群内目标的分布情况,从而提高了群中心的计算稳定性。仿真试验表明,与传统群中心算法相比,IGC-SGT算法提高了目标误差和杂波较大情况下,群目标整体跟踪的可靠性和有效性,并为后续群内单目标的跟踪提供了稳定且可靠的群运动信息。然后,本文针对群内单个目标跟踪困难的问题,提出了一种空间群目标下改进的多假设跟踪方法——GMHT算法。该算法利用近地轨道空间群目标(如弹道导弹中段群目标)运动整体信息,结合多假设跟踪算法的优势,建立了基于空间群目标的假设生成方法、得分计算和剪枝策略,设计了完整的算法流程。仿真试验证明,与传统的多目标关联跟踪算法相比,本文所提算法能根据整体运动趋势自适应调整群内目标与量测的关联,提高单目标的关联精度,减小虚漏警和杂波在跟踪空间密集群目标时的干扰,提高了空间群目标的跟踪性能。最后,本文基于前两章的研究成果,利用MATLAB GUI设计实现了空间群目标数据处理系统,完整解决空间密集群目标的跟踪问题。该系统融合了本文提IGC-SGT算法和GMHT算法,可模拟空间群目标的真实航迹,仿真雷达信号处理中的虚漏警概率情况。实验结果表明,在密集空间群目标下,当目标数目从8增加到16个时,嵌入了IGC-GMHT算法的空间群目标数据处理系统,能够在空间群目标环境中,提取精确稳定的群运动信息,并稳定跟踪群内单个目标,其跟踪误差和稳定性均优于传统多目标跟踪算法。
[Abstract]:Ballistic missiles are usually accompanied by decoys, jammers and launch debris in the middle of the trajectory, forming a large group of threat targets, with small distances and approximately the same trajectory in space. Ground-based radar is the main monitoring method for the middle target. When the traditional multi-target tracking method is used to track the dense space group target, the tracking information error of the individual target is extremely large or even missing due to the limited resources. Therefore, it is very important to study the track tracking of a single target in space group to improve the multi-target tracking capability of ballistic missile defense radar and to analyze the penetration effectiveness of ballistic missile. In this paper, the causes of formation of space group targets in the middle part of trajectory are analyzed, and the motion characteristics of this special dense group targets are briefly described. Then, the common theory of group target tracking is introduced and analyzed. This paper briefly introduces and analyzes the data association algorithm and tracking filtering algorithm for target tracking, which lays a theoretical foundation for the research of multi-hypothesis tracking algorithm under the background space group target. Then, a cluster center tracking method (IGC-SGT) based on cluster clustering is proposed, which improves the stability of group global tracking. Aiming at the problem that the fluctuation and error of the target in the group seriously affect the stability of the cluster center, the IGC-SGT algorithm improves the cluster clustering method, and makes full use of the tracking of the cluster clustering to obtain the distribution of the target in the group. Thus the computational stability of the group center is improved. The simulation results show that compared with the traditional group center algorithm, the IGC-SGT algorithm improves the reliability and effectiveness of the whole tracking of the group target under the condition of large target error and clutter. It also provides stable and reliable group motion information for the tracking of single targets in subsequent groups. Then, aiming at the difficulty of tracking a single target in a group, an improved multi-hypothesis tracking method called GMHT algorithm is proposed. Based on the global motion information of space group targets in low Earth orbit (such as ballistic missile) and the advantages of multi-hypothesis tracking algorithm, a hypothesis generation method, score calculation and pruning strategy based on space group targets are proposed. A complete algorithm flow is designed. The simulation results show that compared with the traditional multi-target association tracking algorithm, the proposed algorithm can adaptively adjust the correlation between the target and the measurement according to the overall motion trend, and improve the correlation accuracy of the single target. The tracking performance of space group targets is improved by reducing the interference of false alarm and clutter in tracking dense group targets. Finally, based on the research results of the first two chapters, this paper designs and implements the spatial group target data processing system by using MATLAB GUI, which solves the problem of tracking the spatial dense group target completely. This system combines IGC-SGT algorithm and GMHT algorithm in this paper. It can simulate the real track of space group targets and simulate the false alarm probability in radar signal processing. The experimental results show that when the number of targets increases from 8 to 16, the spatial group target data processing system embedded with IGC-GMHT algorithm can extract accurate and stable group motion information in the environment of spatial group target. The tracking error and stability of a single target in a stable tracking group are better than those of the traditional multi-target tracking algorithm.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TJ761.3;TN953

【参考文献】