基于神经网络的参差MTI无模糊测速  

Unambiguous Measurement of Velocity for Stagger MTI Pulses Based on Neural Network

作　　者：尹魏昕[1] 于文震[1] 张仕元[1] 

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《微波学报》2010年第6期93-96,共4页Journal of Microwaves

摘　　要：针对MTI(Moving Target Identification,动目标显示)脉冲串线性处理方法速度模糊严重的情况,应用BP(Back Propagation,反向传播)神经网络对参差MTI的速度模糊函数进行优化。通过神经网络的非线性映射能力,使MTI参差脉冲串的多普勒响应接近理想的狄拉克函数,从而解决了传统线性方法对于参差脉冲测速的模糊。仿真结果表明,对于某典型6脉冲MTI回波序列,测速精度可达到5m/s。In view of the serious Doppler ambiguity problem of linear processing of MTI pulses,a new method adopting BP neural network to optimize the Doppler ambiguity function of stagger MTI is proposed.In virtue of the nonlinear mapping ability of neural network,Doppler response of stagger MTI pulses can approach the ideal Dirac function.Thereby the ambiguous problem of stagger pulses velocity measurement can be solved,and a accuracy of 5m/s velocity is gained for a typical 6-pulse MTI sequences.

关 键 词：动目标显示 速度模糊 神经网络 参差脉冲测速 

分 类 号：TN953.1[电子电信—信号与信息处理]