论文部分内容阅读
在自回归谱分析中,Marple的LS算法是最有效的方法之一,它减少了谱线频率的偏移,避免了谱线的分裂,而且具有更高的分辨率.但是,判阶问题仍未解决.本文从实验的角度详细讨论了Marple提出的病态条件和两个容限因子,并且研究了在不同的噪声电平下,分辨率和阶数的关系.最后还讨论了出现假峰的问题.
In autoregressive spectral analysis, Marple’s LS algorithm is one of the most effective methods, which reduces the spectral line frequency offset, to avoid spectral splitting, and has a higher resolution. However, the problem of arbitration Unresolved.This paper discusses the pathological conditions and two tolerance factors proposed by Marple in detail from the experimental point of view, and studies the relationship between resolution and order under different noise levels.Finally, problem.