北京航空航天大学学报 ›› 2019, Vol. 45 ›› Issue (4): 687-694.doi: 10.13700/j.bh.1001-5965.2018.0447

• 论文 • 上一篇    下一篇

基于运动模型的低空非合作无人机目标识别

陈唯实1, 刘佳1,2, 陈小龙3, 李敬1   

  1. 1. 中国民航科学技术研究院, 北京 100028;
    2. 北京航空航天大学 交通科学与工程学院, 北京 100083;
    3. 海军航空大学, 烟台 264001
  • 收稿日期:2018-07-25 出版日期:2019-04-20 发布日期:2019-04-26
  • 通讯作者: 陈唯实 E-mail:chenwsh@mail.castc.org.cn
  • 作者简介:陈唯实,男,高级工程师。主要研究方向:低空空域安全监视、雷达目标检测与跟踪、机场安全运行技术。
  • 基金资助:
    国家自然科学基金委员会-中国民用航空局民航联合研究基金(U1633122);国家重点研发计划(2016YFC0800406)

Non-cooperative UAV target recognition in low-altitude airspace based on motion model

CHEN Weishi1, LIU Jia1,2, CHEN Xiaolong3, LI Jing1   

  1. 1. China Academy of Civil Aviation Science and Technology, Beijing 100028, China;
    2. School of Transportation Science and Engineering, Beihang University, Beijing 100083, China;
    3. Naval Aviation University, Yantai 264001, China
  • Received:2018-07-25 Online:2019-04-20 Published:2019-04-26
  • Supported by:
    Joint Research Foundation of National Natural Science Foundation of China (NSFC) and Civil Aviation Administration of China (CAAC) (U1633122); National Key R & D Program of China (2016YFC0800406)s

摘要: 为保障低空安全,在利用雷达数据探测无人机目标的同时剔除飞鸟等虚警信息,提出了一种基于运动模型的低空非合作无人机目标识别方法,作为已有目标跟踪方法应用的拓展。首先,建立多种运动模型模拟无人机和飞鸟目标运动;然后,基于多种运动模型进行目标跟踪,并估计各种运动模型的出现概率;最后,以各种运动模型在连续时间内出现概率的方差均值来度量目标运动模型的转换频率。通过对仿真数据和机场低空监视雷达实测数据的处理,所提方法能够在杂波环境中跟踪无人机目标并剔除飞鸟目标,进一步验证了其有效性和实用性。

关键词: 运动模型, 无人机, 目标, 识别, 雷达

Abstract: To guarantee the safety of low-altitude airspace, a target recognition method based on motion model was proposed for the non-cooperative UAV target in low-altitude airspace, as an extension of the application of the existing target tracking algorithm, which could detect the UAV and reject the false alarms such as flying birds with radar data. Firstly, multiple motion models were established to simulate the movement of UAV and flying bird targets. Secondly, the targets were tracked with multiple motion models and the appearance probabilities of these models were estimated. Thirdly, the transformation frequency between target motion models was measured by the mean variance of the appearance probabilities of multiple models in continuous time domain. By processing the simulation data and the measured data of the airport low-altitude surveillance radar, the method can track the UAV target in cluttered environment and eliminate the flying bird target, further verifying its effectiveness and practicability.

Key words: motion model, UAV, target, recognition, radar

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发