首页 | 本学科首页   官方微博 | 高级检索  
     

基于神经网络的TWTA非线性校正方法
引用本文:高金雍,董瑶,杜利波,张秀军. 基于神经网络的TWTA非线性校正方法[J]. 河北工业大学学报, 2008, 37(1): 27-32
作者姓名:高金雍  董瑶  杜利波  张秀军
作者单位:河北工业大学信息工程学院,天津,300401;河北工业大学管理学院,天津,300401
基金项目:国家自然科学基金资助项目(60377020;60673087)
摘    要:用于卫星转发器的高功率行波管放大器(TWTA)是非线性放大器,由于常工作在饱和状态导致非线性失真很严重.针对这种非线性变换特性,提出了一种基于神经网络的新的校正方法,主要步骤包括样本数据的预处理、网络结构的设计、采用LM算法进行网络训练,以及利用训练好的模型进行AM/PM或AM/AM校正等.通过仿真实验和对比分析表明,基于神经网络的校正方法优于传统校正方法,校正结果与真实值吻合程度高.

关 键 词:行渡管放大器(TWTA) 非线性校正 神经网络 LM网络
文章编号:1008-2373(2008)01-0027-06
收稿时间:2007-03-30

The Method of Nonlinear Correction of TWTA Based on Neural Network
GAO Jin-yong,DONG Yao,DU Li-bo,ZHANG Xiu-jun. The Method of Nonlinear Correction of TWTA Based on Neural Network[J]. Journal of Hebei University of Technology, 2008, 37(1): 27-32
Authors:GAO Jin-yong  DONG Yao  DU Li-bo  ZHANG Xiu-jun
Affiliation:GAO Jin-yong1,DONG Yao1,DU Li-bo2,ZHANG Xiu-jun1 ( 1. School of Information Engineering,Hebei University of Technology,Tianjin 300401,China,2. School of Management,China )
Abstract:TWTA is a nonlinear amplifier applied in satellite translator, which is susceptible to nonlinear especially on the saturation point. Aiming at the property of nonlinear transformation, a new correction method based on neural network is presented. The main process is to preprocess samples, design the network structure, train the network with LM algorithm, correct AM/PM or AM/AM with the trained network and so on, The simulation experiment and comparison analysis show that the correction method based on LM algorithm is superior to traditional method, and the correction result is consistent exactly with the true value.
Keywords:TWTA    nonlinear correction   neural network   LM network
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号