Analysis of parameter estimation using the sampling-type algorithm of discrete fractional Fourier transform

Parameter estimation is analyzed using two kinds of common sampling-type DFRFT (discrete fractional Fourier transform) algorithm. A model of parameter estimation is established. The factors which influence estimation accuracy are analyzed. And the simulation is made to verify the conclusions. From the theoretic analysis and simulation verification, it can be drawn that, for the estimation of chirp-rate and initial frequency, Pei's method [10] is more suitable if the absolute value of chirp-rate is small relatively; Ozaktas' method [9] is more suitable if the absolute value of chirp-rate is large relatively; and the two methods are both workable if the absolute value of chirp-rate is moderate. The scope of moderate chirp-rate can be approximately determined as [40 Hz/s, 110 Hz/s].     

基于CZT的频谱细化算法及应用

针对复杂信号频域特征提取问题,利用Delphi语言编写了Zoom-FFT和基于CZT的频谱细化方法应用软件,以某型坦克传动箱振动速度信号为例进行了对比分析,验证了算法的有效性和程序设计的正确性。结果表明：基于CZT的频谱细化方法具有运算效率高、细化效果好等优点,可以在复杂信号分析中推广应用。     

基于Chirp变换的快速离散时间尺度变换

提出了一种新的快速离散时间尺度变换算法。给出了离散时间信号尺度变换的构造表达式及其Chirp变换快速实现流程。讨论了Chirp变换实现中的参数选取准则,分析了Chirp变换的信号带宽问题。讨论了算法的运算效率并与其它算法进行了比较。仿真试验证明了本文算法的有效性。