一种基于Padé逼近对短时序列频率和品质因数的估计方法  

SHORT DATA SIGNAL FREQUENCY AND QUALITY FACTOR ESTIMATION ALGORITHM BASED ON PADE APPROXIMANTS

作　　者：曲延锋[1] 国伟华[2] 王守觉[1] 

机构地区：[1]中国科学院半导体研究所神经网络实验室,北京100083 [2]中国科学院半导体研究所集成光电国家重点实验室,北京100083

出　　处：《电子与信息学报》2002年第8期1054-1059,共6页Journal of Electronics & Information Technology

摘　　要：该文提出了一种基于Pade逼近的谱估计方法。由于Pade逼近能够利用有限项很好地逼近对无限项的求和,文中算法适用于短时序列情况。经过实验仿真并与FFT方法和小波变换谱估计方法进行比较表明,该文算法具有较高的频率分辨率和估计精度,并具有较好的品质因数。In this paper a new algorithm for estimating frequency and quality factor based on Fade aproximants is given. Because Fade approximants can accelerate the convergence of sum in the discrete Fourier transform, this algorithm is also used in the frequency and quality factor estimation of short data series. The simulation results show that this algorithm can obtain higher frequency resolution, higher accuracy and better quality factor than FFT and spectrum estimation based on wavelet transformation.

关 键 词：短时序列 PADÉ逼近 FFT 小波变换 频率分辨率 品质因数 谱估计 

分 类 号：TN911.6[电子电信—通信与信息系统]