论文部分内容阅读
在频率分析中使用普通的快速傅立叶变换(FFT)技术时,由其产生的频谱分布在直流到给定的最大频率范围内。频段扩展快速傅立叶变换(ZOOMFFT)提供的是一种从频谱的频率范围的一小部份中获得高分辨率的方法。本文讨论了当今实现频段扩展快速傅立叶变换分析的两种不同的原理。文章指出,这些方法中最重要的区别就在于是否有能力在频段扩展快速傅立叶变换的计算过程中存贮并且保持用于分析的原始时间信号。一旦时间信号能够被存贮,那么这种新的方法就可以用于非平稳信号的分析。文章根据给出的若干频段扩展快速傅立叶变换的典型应用实例对这些方法做了比较。
When using conventional Fast Fourier Transform (FFT) techniques in frequency analysis, the resulting spectrum is distributed from dc to a given maximum frequency range. Band Expansion Fast Fourier Transform (ZOOMFFT) provides a way to get high resolution from a small portion of the frequency spectrum of a spectrum. This article discusses two different principles for implementing fast Fourier transform analysis of the band today. The article points out that the most important difference among these methods is the ability to store and maintain the original time signal for analysis during the calculation of the band-extended Fast Fourier Transform. Once the time signal can be stored, then this new method can be used for non-stationary signal analysis. The paper compares these methods according to the typical application example of extended Fast Fourier Transform for several frequency bands.