针对传统基于雷达散射截面积(RCS)的微动特征提取验证所使用数据源精度不高、未考虑弹头平动等问题,文中搭建了分布式特征提取验证系统,对常用进动周期提取方法进行了验证。首先,分别使用STK和FEKO软件生成了弹头的运动特性及弹头RCS模板;其次,将运动特性同RCS模板进行复合,生成中段弹头的高精度RCS序列;最后,基于生成序列,验证常见算法的提取效果。仿真验证表明:提取方法的正确率不仅受制于噪声强度,并且同平均视界角的取值密切相关,当平均视界角小于35毅时,自相关函数法相对多重自相关方法正确率高出2倍以上。

For the problems of the traditional method of micro feature extraction, such as low accuracy of data source, not considering the warhead translatory motion, the distributed authentication system of feature extraction is built and the commonly used precession period extraction method is verified. Firstly, the motion characteristic of the middle warhead and warhead radar cross section (RCS) template is generated by using the software of STK and FEKO. Secondly, the high-precision RCS sequence of warhead during middle course is generated by compositing the movement characteristics and the RCS template. Finally, the extraction effect of the common algorithm is verified based on the generated sequence. The simulation shows that the accuracy of pecession period extration method is not only subject to noise intensity but also closely related to the average aspect angle,the ACF algorithm gains more than two times accuracy than the CAUTOC algorithm when the average aspect angle is less than 35°.

TJ761.3

国家自然科学基金资助项目