首页 | 本学科首页   官方微博 | 高级检索  
     

基于SVD-TLS的ARMA谱估计及应用
引用本文:朱学锋,吴晓琳.基于SVD-TLS的ARMA谱估计及应用[J].战术导弹技术,2004(4):M029-33.
作者姓名:朱学锋  吴晓琳
作者单位:92941部队96分队,辽宁葫芦岛125001
摘    要:提出了基于SVD-TLS法估计ARMA信号模型的AR阶数和参数的方法, 并利用Cadzow谱估计子计算ARMA功率谱.在对包含两个非常接近的频率成份的模拟数据进行计算机仿真的基础上, 提出了利用SVD-TLS法计算ARMA功率谱存在的问题和建议.对真实的动态观测数据计算ARMA谱, 并与Burg最大熵谱法和传统的周期图法进行了对比分析.计算机仿真结果表明, SVD-TLS法可以用较少的AR阶数实现高分辨率的ARMA功率谱估计.

关 键 词:奇异值法  总体最小二乘法  ARMA功率谱
文章编号:1009-1300(2004)04-0029-05
修稿时间:2004年4月27日

Spectral Analysis Using ARMA Models Based on SVD-TLS Algorithm
Zhu Xuefeng,Wu Xiaolin.Spectral Analysis Using ARMA Models Based on SVD-TLS Algorithm[J].Tactical Missile Technology,2004(4):M029-33.
Authors:Zhu Xuefeng  Wu Xiaolin
Abstract:An approach to determine the AR order and parameters of autoregressive moving average(ARMA) model by singular value decomposition(SVD) and total least square(TLS)is present, and ARMA power spectrum is calculated with Cadzow spectral estimator.With the statistical result provided by computer simulation about simulated data sequence with two close frequencies,the problem existing in computing ARMA spectrum estimation is point out by the method,and some advice is given.The application in real observed data estimates the ARMA power spectrum by the proposed algorithm,demonstrated by comparing with Burg maximum entropy approach.Computer simulation shows that the estimated spectrum with SVD TLS algorithm provides better resolution than with Burg algorithm and conventional processing by less AR order.
Keywords:singular value decomposition  total least square  ARMA power spectrum
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