【摘要】 首次将时间重叠法和ssinc窗应用于循环谱估计中,提出了时间重叠的循环谱估计方法,使循环谱估计取得了突破性的进展(:1)循环谱的偏离误差减小,有效频率分辨率提高,能准确地分辨出循环频率方向和谱频率轴方向的谱峰;(2)计算量大大减少。通过时间的重叠过程,减少DFT要求的数据量,提高了循环谱估计的实时性。更多还原

【Abstract】 This article presents a new method of cyclic spectra estimation which uses time aliasing method and ssinc window in it.It is the first time that the time aliasing method and ssinc window are adopted in the history of the estimation of cyclic spectra.This method makes great progresses as follows in the estimation of cyclic spectra:First,the quality of the cyclic spectra is raised excellently by the improvement of effective frequency resolution and the reduction of bias error,and the spectrum peaks can be identified clearly both in cyclic frequency direction and in the spectrum frequency direction.Second,the burden of calculation is lessened greatly by the time aliasing process.It reduces the quantity of data in DFT,and increases the capability for the real time processing in cyclic spectra estimation.更多还原