HHT的改进及其在时频信号分析中的应用研究  

An Improved HHT and Its Application in Time-Frequency Signals Analysis

作　　者：耿婷婷[1] 金荣洪[1] 耿军平[1] 梁仙灵[1] 

机构地区：[1]上海交通大学电子工程系,上海200240

出　　处：《中国电子科学研究院学报》2010年第5期513-517,共5页Journal of China Academy of Electronics and Information Technology

基　　金：国家自然科学基金委创新研究群体基金项目(60821062);"973"项目(2009CB320205)

摘　　要：在希尔伯特—黄变换的筛选过程中,用三次样条插值法求取极值包络通常会出现"过冲"问题。根据本征模态分解原理,提出了极值序列加密方法,并通过与极值延拓法结合,得到改进的HHT,该算法可以较好的改善"过冲"问题,得到更为准确的信号特征。应用改进的HHT方法分析了线性调频连续波(FMCW)信号,仿真结果表明了改进方法对于低信噪比条件下线性调频信号的提取更彻底,参数估计更准确。In the shifting process of Hilbert-Huang transform,there is usually an ＂overshoot and undershoot＂ problem by using cubic spline interpolation method for the extremum envelopes.Based on the theory of Empirical Mode Decomposition,a method of increasing the extreme points into the extreme sequence is proposed.Combining with the boundary endpoint extension method,this method can improve the ＂overshoot and undershoot＂ problem and get more accurate signal characteristics.For frequency modulation continuous wave,the simulation results show that the improved HHT method can extract more characteristics for the low SNR non-stationary signals than the conventional HHT method,and the parameter estimations are more accurate.

关 键 词：本征模态分解 希尔伯特—黄变换 拟合过冲 极值加密与延拓 

分 类 号：TN957.51[电子电信—信号与信息处理]