论文部分内容阅读
1979年美国J·E·Shore首次提出最小交叉熵谱分析法。理论证明,最大熵谱分析法只是最小交叉熵法的一个特例。当存在先验信息时,最小交叉熵法可获得比最大熵法好得多的分辨率。因此最小交叉熵法一出现,就受到广泛注意。但几年来它的进展不快,原因之一是运算太繁复,并且Shore等人至今仍未公布其具体算法。本文概括地介绍了最小交叉熵法发展概况,以及国内外学者提出的几种主要方法(包括李建三硕士和作者提出的快速倒谱算法,以及“熵谱分析程序包ESAP”)。并通过实际运算评述了各种方法的优缺点。文末指出,尽管目前对最小交叉熵法的研究尚不充分,但这是一种很有发展前途的谱分析法。
In 1979, JE Shore first proposed minimum cross-entropy spectral analysis. Theory shows that the maximum entropy spectral analysis is only a special case of the minimum cross entropy method. When there is a priori information, the minimum cross-entropy method can get a much better resolution than the maximum entropy method. Therefore, the minimum cross-entropy method appears, it received widespread attention. But one of the reasons its progress has been unpleasant for years is that computation is too complicated, and so far, Shore et al. Have not yet published their specific algorithm. This article provides an overview of the development of minimum cross entropy method, and several main methods proposed by domestic and foreign scholars (including fast cepstrum algorithm proposed by Li Jiansan and authors, and “Entropy Spectrum Analysis Package ESAP”). Through the actual operation, the advantages and disadvantages of various methods are commented. At the end of the article, it is pointed out that although there is not enough research on the minimum cross entropy method at present, this is a very promising spectral analysis method.