一种基于改进Hilbert-Huang变换的非平稳信号时频分析法及其应用  

A Time-Frequency Analysis Method for Non-stationary Signals Based on Improved Hilbert-Huang Transform and Its Application

作　　者：毛炜[1] 金荣洪[1] 耿军平[1] 李家强[1] 

机构地区：[1]上海交通大学电子工程系,上海200240

出　　处：《上海交通大学学报》2006年第5期724-729,共6页Journal of Shanghai Jiaotong University

基　　金：航天支撑技术基金资助项目(2002-HT-SHJD)

摘　　要：对于非平稳信号时频分析,提出了一种基于改进Hilbert-Huang变换(HHT)的分析方法.根据HHT的已有原理,改进了经验模式分解(EMD)过程中的筛选停止准则,提高了分解精度;给出了Hilbert谱分析的完整过程;以线性调频连续波(FMCW)信号模型作为研究对象,结合改进的EMD分解和完整的Hilbert谱分析,通过分析时频分布特征实现高噪声背景下雷达目标信号的检测以及干扰信号的提取.仿真结果表明了改进后的HHT方法对于低信噪比非平稳信号分析的有效性.A time frequency analysis method for non-stationary signals, deriving from the improved Hilr bert-Huang Transform （HHT）, was proposed. Based on the current theory of HHT, the sifting stop guide is improved so that the precision and correctness of the empirical mode decomposition （EMD） is enhanced. The whole process of Hilbert spectrum analysis is also given and is combined with the corresponding modified EMD. Through the time-frequency distribution characteristics, the HHT method is adopted to detect radar target signal in the form of linear frequency modulated continuous waveform （FMCW） signals in strong noise background and extract the interference one as important potential applications. The simulation results suggest the effectiveness of this new method in low-SNR non-stationary signal analysis.

关 键 词：非平稳信号 低信噪比 瞬时频率 Hilbert—Huang变换 雷达 

分 类 号：TN957.51[电子电信—信号与信息处理]