论文部分内容阅读
本文提出两个新的功率谱估值方法——双最大熵迭代法及模拟噪声迭代法。它们均出发于比最大熵估值法更一般的信号模式:自回归——滑动平均模式,扩展了估值适应的范围,提高了估值准确度。它们都具有比最大熵估值法更好的估值特性,并只要求低得多的估值阶数。这两方法也继承了最大熵估值法的优点,能很好地处理短输入数据。它们适用于要求准确估测功率谱分布的场合。本文叙述了推导过程,并列举了部分计算机模拟试验结果。
In this paper, two new power spectral estimation methods are proposed-double maximum entropy iteration and analog noise iteration. They all originate from a more general signal model than the maximum entropy method: the autoregressive-moving averaging model, which expands the range of valuation adaptation and improves the accuracy of the estimation. They all have better valuation than the maximum entropy method and require only much lower order of valuation. These two methods also inherit the advantages of the maximum entropy estimation method, which can well deal with short input data. They are suitable for applications that require accurate estimation of the power spectrum distribution. This article describes the derivation process and lists some of the computer simulation results.