线性调频Z变换在信号频谱分析中的应用

由于快速傅里叶变换（Fast Fourier Transform,FFT）算法不能精确反应信号的局部频谱特性,对此,本文以按时间抽取（DIT）的基-2FFT算法为基础,并参考基于FFT的布鲁斯坦（Bluestein）算法,设计了新的信号频谱分析软件,用于对实序列采样信号做线性调频Z变换,即频率抽样处在Z平面上,可沿任意螺线做频率抽样的频谱分析方法.结合工程实践对相同采样点数的信号在0—50Hz频率段做频谱分析,由频谱图可以看到,采用线性调频Z变换算法远比采用FFT变换算法求得信号的频谱更精确.

Chirp-Z Transform and its Applications in Spectrum Analysis

FFT(fast fourier transform) algorithm can not present the characteristic of local spectrum precisely. Regarding this, in this paper was designed a new software for spectrum analysis, based on radix 2 based decimation in time FFT and Bluestein FFT algorithm, which realized chirp-z transform on real sequence sampling signa. The frequency spectrum analysis method that can frequency sample along the arbitrary spire in the Z-plane. in practice we do spectrum analysis to the same signal at the sampling points of 0-50Hz frequency bands, from the chart of spectrum. Results show that the Signal spectrum using Chirp Z Transform Algorithm is more accurate compared with using Transform FFT algorithm.