运用分频段希尔伯特黄变换进行多分量信号的频散分析  

Analysis of Frequency Dispersion of Multi-component Signal using Sub-band Hilbert-Huang Transform

作　　者：蒋礼[1,2] 

机构地区：[1]中国地质大学(武汉)地球物理与空间信息学院,武汉430074 [2]华北水利水电学院数学与信息科学学院,郑州450011

出　　处：《电讯技术》2012年第4期472-477,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(40974079)~~

摘　　要：将分频段希尔伯特黄变换应用于多分量信号的频散分析中。首先,利用带通滤波器和经验模态分解相结合,成功实现了经验模态频率分解,并准确提取了经验频率模态函数;然后,使用该方法准确地获取了多分量含噪信号的时频能量谱和时频相位谱;最后,基于同步相差和异步相差算法,精确绘制了原信号的相速度频散分析曲线。数值试验表明,该算法拥有较高的时频分辨能力和良好的抗噪性能,对于复杂且信噪比较低的信号,能获得比传统希尔伯特黄变换更准确的频散分析结果。In this paper, the sub - band Hilbert - Huang transform （S- HHT） is used to analyse the frequency dispersion of multi - component signal. Firstly, the band - pass filter and the empirical mode decomposition are combined to achieve the empirical mode frequency decomposition （EMFD） successfully, and accurately obtain the empirical frequency mode functions （EFMF）. Secondly, the time frequency energy spectrums and time frequency phase spectrums from multi - component signal with noise are precisely drawn by the method. Finally, by using the synchronous phase difference and asynchronous phase difference arithmetic the phase velocity disper- sion curves are drawn accurately. The numerical results show that the S - HHT arithmetic has high time frequen- cy resolution and good resistance to noise. For complex and low Signal- to- Noise Ratio（SNR） signal, the S - HHT can get the more accurate results of dispersion analysis than the HHT.

关 键 词：希尔伯特黄变换 经验模态频率分解 经验频率模态函数 瞬时频率 频散曲线 

分 类 号：TN911[电子电信—通信与信息系统]