广东工业大学学报 ›› 2020, Vol. 37 ›› Issue (02): 67-73.doi: 10.12052/gdutxb.190074

• 综合研究 • 上一篇    下一篇

基于循环神经网络的目标转弯机动类型识别

吴家湖1, 熊华2, 宗睿2, 赵曜1, 周贤中1   

  1. 1. 广东工业大学 信息工程学院, 广东 广州 510006;
    2. 北京电子工程总体研究所, 北京 100854
  • 收稿日期:2019-05-25 出版日期:2020-03-10 发布日期:2020-01-13
  • 通信作者: 赵曜(1984-),男,副教授,硕士生导师,主要研究方向为雷达系统设计、合成孔径雷达信号处理、雷达跟踪滤波算法、稀疏信号处理等,E-mail:zhaoyao431@163.com E-mail:zhaoyao431@163.com
  • 作者简介:吴家湖(1993-),男,硕士研究生,主要研究方向为雷达跟踪滤波算法、图像处理、机器学习等
  • 基金资助:
    国家自然科学基金青年基金资助项目（61704032）；中国科学院计算技术研究所计算机体系结构国家重点实验室开放课题（CARCH201814）

Target Turning Maneuver Type Recognition Based on Recurrent Neural Networks

Wu Jia-hu1, Xiong Hua2, Zong Rui2, Zhao Yao1, Zhou Xian-zhong1   

  1. 1. School of Information Engineering, Guangdong University of Technology, Guangzhou 510006, China;
    2. Beijing Institute of Electronic System Engineering, Beijing 100854, China
  • Received:2019-05-25 Online:2020-03-10 Published:2020-01-13

HTML

PDF (PC)

2786

摘要： 对战斗机等空中目标机动类型进行识别是掌握其战术意图的重要依据。为了更好地识别对方战机的机动类型，提升作战能力，本文主要研究空中高速运动目标（如战斗机）的机动类型识别，针对传统的机动类型识别算法的识别准确率不高的问题，将循环神经网络运用在机动类型的识别上，利用Bi-LSTM、LSTM、RNN和GRU循环神经网络识别转弯机动类型。实验结果表明，循环神经网络能够高效识别目标的转弯机动类型，Bi-LSTM的识别准确率达到了98.85%。

关键词: 机动目标, 机动类型, 循环神经网络, 双向长短期记忆网络, 机动识别

Abstract: Recognizing the maneuver types of airborne objects, such as fighters, is an important identification basis in figuring out their military intentions. To better identify the performance of maneuver types of enemy warfare aircraft and thus enhance our military's combat capability, the airborne high-speed moving targets such as the fighter's maneuver types are mainly studied. Addressing the low identification accuracy of traditional methods, recurrent neural networks are adopted to the problem of maneuver types recognition and identification. The main contribution of this study is the application of four kinds of recurrent neural networks architectures:Bi-LSTM (Bidirectional LSTM), LSTM, RNN and GRU. The experiment results show that the proposed RNN based method can recognize the target's turning maneuver type with efficiency, and the identification accuracy of Bi-LSTM is achieved by 98.85%.

Key words: maneuvering target, maneuver type, recurrent neural networks, Bi-LSTM (Bidirectional LSTM), maneuver recognition

中图分类号: 

  • TP183
[1] 冷画屏, 吴晓锋, 王慕鸿. 空中目标战术机动类型的实时识别[J]. 火力与指挥控制, 2011, 36(1):64-66 LENG H P, WU X F, WANG M H. Real-time recognition of air target tactical maneuvering type[J]. Fire Control & Command Control, 2011, 36(1):64-66
[2] GERDES A. Automatic maneuver recognition in the automobile:the fusion of uncertain sensor values using Bayesian models[C]//Proceedings of the 3rdInternational Workshop on Intelligent Transportation. Hamburg, Deutschland:WIT, 2006, 129-133.
[3] HUENHAGEN T, DENGLER I, TAMKE A, et al. Maneuver recognition using probabilistic finite-state machines and fuzzy logic[C]//Proceedings of the IEEE Conference on Intelligent Vehicles Symp. (IV), La Jolla, CA:IEEE, 2010, 65-70.
[4] TKACHENKO P, ZHOU J, GAGLIARDI D, et al. On-line maneuver identification in highway traffic using elastic template matching[J]. IFAC-Papers on Line, 2018, 51(15):557-562
[5] 杨晓伟, 赵海军. 自动目标识别与机动目标建模及跟踪方法研究[J]. 电子科技, 2017, 30(10):127-130 YANG X W, ZHAO H J. Research on automatic target recognition and maneuvering target modeling and tracking[J]. Electronic Sci & Tech, 2017, 30(10):127-130
[6] MILAN A, REZATOFIGHI S H, DICK A, et al. Online multi-target tracking using recurrent neural networks[C]//Proceedings of the AAAI conference on Artificial Intelligence. San Francisco, California:AAAI, 2017.
[7] LI X R, JILKOV V P. Survey of maneuvering target tracking. Part I. Dynamic models[J]. IEEE Transactions on Aerospace & Electronic Systems, 2004, 39(4):1333-1364
[8] RICHARDS M A, SCHEER J A, HOLM W A. Principle of Modern Radar:Basic Principles[M]. Rleigh, NC:SciTech, Publishing, 2010.
[9] 杨丽, 吴雨茜, 王俊丽, 刘义理. 循环神经网络研究综述[J]. 计算机应用, 2018, 38(S2):1-6, 26 YANG L, WU Y X, WANG J L, et al. Research on recurrent neural network[J]. Journal of Computer Applications, 2018, 38(S2):1-6, 26
[10] LIPTON Z C, BERKOWITZ J, ELKAN C. A critical review of recurrent neural networks for sequence learning[J]. Computer Science, 2015, 10:1-38
[11] 张荣, 李伟平, 莫同. 深度学习研究综述[J]. 信息与控制, 2018, 47(4):385-397 ZAHNG R, LI W P, MO T. Review of deep learning[J]. Information and Control, 2018, 47(4):385-397
[12] GREFF K, SRIVASTAVA R K, KOUTNIK J, et al. LSTM:A search space odyssey[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(10):2222-2232
[13] GRAVES A, FERNANDEZ S, SCHMI-DHUBER J. Bidirectional LSTM networks for improved phoneme classification and recognition[C]//Proceedings of the International Conference on Artificial Neural Networks (ICANN). Berlin, Heidelberg:Springer, 2005:799-804.
[14] CHUNG J, GALCEHRE C, CHO K H, et al. Empirical evaluation of gated recurrent neural networks on sequence modeling[J]. Eprint Arxiv, 2014, 11:1-10
[15] 何友, 修建娟, 关欣. 雷达数据处理及应用[M]. 北京:电子工业出版社, 2013:60-70.
[16] 周宏仁, 敬忠良, 王培德. 机动目标跟踪[M]. 北京:国防工业出版, 1991:78-108.
[1] 张舒, 莫赞, 柳建华, 杨培琛, 刘洪伟. 基于NWD集成算法的多粒度微博用户兴趣画像构建[J]. 广东工业大学学报, 2020, 37(04): 42-50.
[2] 张文峰； 胡振涛； 程建兴； . 种车辆机动目标跟踪的多传感器信息融合估计算法[J]. 广东工业大学学报, 2009, 26(1): 16-.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发