论文部分内容阅读
把离散傅里叶变换DFT改为快速傅里叶变换FFT,可以大大节约用机的时间和内存。但是,应用FFT进行傅里叶变换,要满足离散序列的点数n=2~p,p=1,2,…,m,即n=2,4,8,…,2~p。DFT则不需要满足离散序列的点数n=2~p。假设我们对某一时窗内的时间序列进行傅里叶变换,设时窗长度为l,于是,离散序列点数为n=l/Δt,Δt是采样间隔。如果这时的n=2~p,那么毫无问题,就可以直接
The discrete Fourier transform DFT to Fast Fourier Transform FFT, you can save time and memory. However, the application of FFT to Fourier transform, to meet the discrete sequence of points n = 2 ~ p, p = 1,2, ..., m, that is n = 2,4,8, ..., 2 ~ p. DFT does not need to meet the discrete sequence of points n = 2 ~ p. Suppose we Fourier transform the time series in a time window, and set the length of the time window to 1, so the discrete sequence points are n = l / Δt and Δt is the sampling interval. If this time n = 2 ~ p, then there is no problem, you can directly