《计算机应用研究》|Application Research of Computers

面向量测的<i>m</i>-最优<i>N</i>扫描多假设跟踪方法

Measurement-oriented <i>m</i>-optimal hypothesis <i>N</i>-scan multi-hypothesis tracking algorithm

免费全文下载 (已被下载 次)  
获取PDF全文
作者 刘建锋
机构 南京航空航天大学 电子信息工程学院,南京 211106
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2021)01-057-0282-05
DOI 10.19734/j.issn.1001-3695.2019.09.0524
摘要 为解决联合概率数据互联只能跟踪已知数目目标和互联模糊,以及传统多假设跟踪算法生成假设数目随时间积累呈指数增长问题。提出一种面向量测<i>m</i>-最优假设<i>N</i>扫描MHT方法。首先,在每一假设下生成<i>m</i>-最优假设,在每帧产生既定数目最优及次优假设;然后,通过<i>N</i>宽度滑窗产生最优可行假设,完成数据互联,并分别使用两点差分线性法和全局最小二乘估计完成单个新目标和多个新目标航迹起始。仿真结果表明,该方法与MHT-DAM算法相比较,获得了跟踪性能和运算时间上的平衡。
关键词 多假设跟踪; 多目标跟踪; 新目标航迹起始; 全局最小二乘估计; 两点差分线性法
基金项目 中国航空基金资助项目(2017ZC52036,20172752019)
南京航空航天大学雷达成像与微波光子学教育部重点实验室资助项目
本文URL http://www.arocmag.com/article/01-2021-01-057.html
英文标题 Measurement-oriented <i>m</i>-optimal hypothesis <i>N</i>-scan multi-hypothesis tracking algorithm
作者英文名 Liu Jianfeng
机构英文名 College of Electronic Information & Engineering,Nanjing University of Aeronautics & Astronautics,Nanjing 211106,China
英文摘要 In view of the problem that the JPDA algorithm can only track the known number of targets and association ambiguity, simultaneously, the number of hypotheses in MHT algorithm increases exponentially over time. This paper proposed a measurement-oriented <i>m</i>-optimal hypothesis <i>N</i>-scan MHT algorithm. Firstly, it used the method of generating m-optimal hypothesis to generate a specified number of optimal and sub-optimal hypothesis in each frame, and used fixed-value <i>N</i> of hypothesis tree depth to generate optimal feasible hypothesis to complete data association. This paper used two-point difference linear method and global least squares estimation to solve the track initiation of single new target and multiple new targets, respectively. The simulation results show that, comparing with MHT-DAM algorithm, the proposed method achieves a balance between tracking performance and operation time.
英文关键词 multi-hypothesis tracking; multi-target tracking; track initiation of new targets; global least squares estimation; two-point difference linear method
参考文献 查看稿件参考文献
 
收稿日期 2019/9/1
修回日期 2019/10/24
页码 282-286,292
中图分类号 TP301
文献标志码 A