基于时频分布的弹道导弹目标识别方法  

Identification Method of Ballistic Missile Target Based on Time-Frequency Distribution

作　　者：李陆军[1,2] 吕金建[1] 丁建江[1] 许红波[1] 刘宇驰[1] 

机构地区：[1]空军预警学院 [2]解放军93975部队

出　　处：《火力与指挥控制》2017年第6期45-50,共6页Fire Control & Command Control

基　　金：国家自然科学基金青年项目(61401503);中国博士后基金(20110491889);全军军事类研究生基金(2014JY545);学院创新基金(2013ZDJC0101);学院创新基金资助项目(2014QNCX0115)

摘　　要：在弹道导弹目标识别中,微动特征是重要的识别手段。从弹道导弹微动特性时频分析出发,提出一种基于时频分布的弹道导弹目标识别方法。该方法将时频分布图的伪Zeinike不变矩特征作为识别特征。首先对回波信号进行时频变换以获取时频图像;然后为了降低噪声的影响,对其进行图形预处理;最后给出了伪Zernike不变矩提取步骤及识别特征的选取原则。通过仿真实验,分析了不同特征组合对识别率的影响,评估了不同信噪比下识别方法的稳定性。实验结果表明,该方法具有一定稳定性,可用于弹道导弹目标识别。Micro-motion feature is an important means in ballistic missile identification.Identification method of ballistic missile target based on time-frequency distribution is proposed in the article,according to time-frequency analysis of ballistic missile target micro-motion. It identifies target mainly by extracting pseudo-zernike invariant moment of time-frequency image as identification feature. In the method,Firstly,time-frequency analysis is applied to get a series of time-frequency images. Then these images are pre-processed to filter noise as two-dimension images. At last,it is given for the process of pseudo-zernike invariant moment extraction and the principle of identification feature selection. By simulation and experiment,whether different feature combinations impact the identification rate is analyzed and the method robustness with different SNR is evaluated. The result shows that the method is stable. It has certain application value in ballistic missile target identification.

关 键 词：弹道导弹 微动特征 时频分布 目标识别 伪Zernike不变矩 

分 类 号：TJ761[兵器科学与技术—武器系统与运用工程]