摘要
最大熵谱分析方法是一种非线性的新的谱分析方法。其基本原理可浅近地解释为:对一个所考虑的时间系列,它的前 m 个自相关系数已知,要求构造其后的自相关系数,但不损失系列的熵即不损失其信息。实际的处理过程是:用自回归方法在最小二乘原则下对系列作最佳拟合,然后求拟合系列的理论谱。最大熵谱法不要求对原始系列的滤波和对粗谱的平滑整理。本文从实际应用的角度,简述了该方法的过程,介绍了编程计算的方法和步骤,并结合实例计算了几组大气湍流谱。文章认为最大熵谱分析方法是一种值得推广应用的优秀的谱分析方法。
The maximum entropy spectral estimation is a new nonlin-ear method of spectral analysis.The basic principle can be explainedas follows:For a weak stationary time series with zero mean,if theformer M values of autocorrelative function are given,the problem isto determine the latter and keep the entropy or the information of theseries unvarying.The real procedure of spectral estimation is to,according to the least square rule,make up an optimal time serieswith the most approachability to the original by means of the auto-regressive method,then estimate its theoretic spectral.The maxiumentropyspectral estimation does not demand the filter of the series or thesmoothing of the raw spectral.The application of the basic formulasabout maximum entropy spectral estimation are discussed in brief,thecalculation method and computer programme are explained in detail,andsome examples of the spectral e estimation are given in this paper.
关键词
傅立叶谱
自回归
熵
湍流
fourier spectrum
autoregression
predication error
entropy
