摘要
针对高轨卫星连线干涉测量(Connected Element Interferometry,CEI)信号的高精度频率估计这一难题,建立了CEI中的正弦信号频率估计模型。设计了基于深度学习框架的CEI信号频率估计算法,将算法划分为基于前馈深度神经网络的频率表征模块和基于卷积神经网络的频率计算及估计模块,在此基础上设计了各模块的具体结构和学习训练流程。对于算法的核心模块进行了仿真实验验证,并将所提算法与前人的相关算法进行了比较与分析,证明了该算法的有效性、稳定性和优越性。
For the difficult problem of high-precision frequency estimation of connected element interferometry(CEI)signals of high-orbit satellites,a frequency estimation model of sinusoidal signals in CEI is established.A CEI frequency estimation algorithm based on deep learning framework is designed.The improved algorithm is divided into frequency representation module based on feedforward deep neural network and frequency calculation and estimation module based on convolutional neural network.The concrete structure as well as learning and training process of each module are designed.The core module of the algorithm is verified by simulation experiments.On this basis,the proposed algorithm is compared with existing algorithms,and the effectiveness,stability and superiority of the algorithm are proved.
作者
高泽夫
杨文革
焦义文
毛飞龙
王育欣
李雪健
张书雅
李超
滕飞
GAO Zefu;YANG Wenge;JIAO Yiwen;MAO Feilong;WANG Yuxin;LI Xuejian;ZHANG Shuya;LI Chao;TENG Fei(Department of Electronic and Optical Engineering,Space Engineering University,Beijing 101416,China;School of Space Command,Space Engineering University,Beijing 101416,China;State Key Laboratory of Complex Electromagnetic Environment Effects on Electronics and Information System,Luoyang 471003,China)
出处
《电讯技术》
北大核心
2023年第11期1687-1695,共9页
Telecommunication Engineering
基金
北京市科技重大专项(Z181100002918004)。
关键词
高轨卫星
连线干涉测量
正弦信号频率估计
深度学习
high-orbit satellite
connected element interferometry(CEI)
frequency estimation of sinusoidal signal
deep learning
