基于多普勒特性的飞机尾流回波提取方法  

Extraction of Aircraft Wake Vortices Radar Returns Based on Doppler Characteristics

作　　者：刘俊凯[1] 李文臣[1] 王雪松[1] 王涛[1] 

机构地区：[1]国防科技大学电子科学与工程学院综合电子信息系统研究所,长沙410073

出　　处：《系统仿真学报》2011年第7期1323-1328,共6页Journal of System Simulation

基　　金：探索研究项目(7130620)

摘　　要：飞机尾流探测技术是航空安全领域的重要研究课题。使用一组均匀分布的粒子模拟尾流涡旋的运动特征,研究了潮湿大气中尾涡多普勒谱的视频回波仿真方法。尾流回波的信噪比很低,降低检测门限的方法导致产生大量虚警。为了克服这个缺点,基于尾涡多普勒特性,提出了一种利用多层前馈神经网络提取尾流回波的方法。仿真表明,尾流回波的多普勒谱谱宽较大,谱型不同于气象回波的高斯谱,通过对比分析验证了神经网络提取方法的有效性。研究结果有助于利用脉冲多普勒雷达对飞机尾流探测技术的研究。Detection of wingtip generated wake vorteces is an important field for aviation safety.Wake vortex movement characteristic was modeled using uniform distributed particles,and a video frequency simulation method was studied for the Doppler Spectral signature of wake vortex in wet air.The wake vortex signal-noise-ratio is low,and the approach of lowering the detection threshold often leads to the generation of many false alarms in the observation space.In order to overcome the shortcoming,based on the wake vortex Doppler characteristics,a wake vortex return extraction method was proposed using a multi-layer feedforward neural network.Simulation results indicate that the Doppler Spectrums have larger spectrum width,and the spectrum shapes differ from ordinary weather signatures of Gauss spectrums.Then the validity of the neural network extraction method was proved through comparison with the conventional method.The research is helpful to studying the detection technology of wake vortex using the pulsed-Doppler radar.

关 键 词：尾流 潮湿大气 多普勒谱 提取 前馈神经网络 

分 类 号：TN957.51[电子电信—信号与信息处理] V211.41[电子电信—信息与通信工程]