只估计AR参数的ARMA(p,q)谱估计的新算法
A New Algorithm for Arma (p,q) Spectral Estimation Using Only AR Parameters Estimation
-
摘要: 本文提出一种估计自回归AR参数的新算法.新算法采用递推Householder变换算法. 文中给出了ARMA(4,4)仿真计算例子及两个正弦加白噪声的仿真计算结果,并与最小二乘 法的计算结果进行了比较.结果表明新算法在分辨率和估计质量方面均优于最小二乘法和已 有的谱估计方法,也说明用提高算法稳定性的方法可解决负谱问题和提高谱估计质量.Abstract: A new algorithm for autoregressive AR parameters estimation is presented in this paper. The recursive Householder algorithm is used in this new algorithm. Simulation results of ARMA (4,4) and two sinusoids with white noise by using both the new method and the traditional LS are given at the same time for comparison. These results show that the resolution and the quality of estimation by this method are better than that of LS method and the existing method. They also show that negative spectrum can be eliminated and the quality of spectrum estimation can also be improved by high stability algorithm.
-
点击查看大图
计量
- 文章访问数: 1431
- HTML全文浏览量: 68
- PDF下载量: 1144
- 被引次数: 0
下载:
