基于高斯混合概率假设密度滤波器的扩展目标跟踪算法

发布时间：2018-03-16 20:09

本文选题：目标跟踪　切入点：多扩展目标　出处：《系统工程与电子技术》2017年03期 　论文类型：期刊论文

【摘要】：针对杂波环境下多扩展目标跟踪中航迹起始和量测集划分问题,提出了一种基于高斯混合概率假设密度滤波器的扩展目标跟踪算法。在航迹起始阶段利用最近邻指数法对量测集进行聚类趋势分析,接着通过改进OPTICS(ordering points to identify the clustering structure)算法,建立一个增广数据集排序来表示量测集的密度结构,该算法对参数选择、初始点选择均不敏感,可以滤除量测集中的杂波。仿真结果表明,在航迹起始阶段本文所提算法在保证起始性能的同时计算代价明显减少,在量测集划分过程中,所提算法能够有效划分不同形状、密度的扩展目标,自适应地确定划分数目,减少算法运行时间。
[Abstract]:To solve the problem of track initiation and measurement set partition in multi-extended target tracking in clutter environment, An extended target tracking algorithm based on Gao Si hybrid probability hypothetical density filter is proposed in this paper. The nearest neighbor exponent method is used to analyze the clustering trend of the measurement set in the initial phase of the track, and then the improved OPTICS(ordering points to identify the clustering structure algorithm is adopted. An extended data set sort is established to represent the density structure of the measurement set. The algorithm is insensitive to parameter selection and initial point selection, and can filter the clutter of the measurement set. The simulation results show that the proposed algorithm is not sensitive to parameter selection and initial point selection. In the phase of track initiation, the algorithm proposed in this paper can reduce the computational cost while ensuring the initial performance. In the course of measurement set partition, the proposed algorithm can effectively partition the extended targets with different shapes and densities, and adaptively determine the number of partitions. Reduce the running time of the algorithm.
【作者单位】：空军工程大学信息与导航学院;空军大连通信士官学校;
【基金】：国家自然科学基金(61571458)资助课题
【分类号】：TN713

【相似文献】