改进EEMD-小波阈值的信号处理方法  

An improved EEMD wavelet-threshold signal processing method

作　　者：康安康[1] KANG An-kang(Unit 91404 of PLA, Qinhuangdao 066000, China)

机构地区：[1]91404部队,河北秦皇岛066000

出　　处：《雷达与对抗》2021年第3期24-27,共4页Radar & ECM

摘　　要：为解决EEMD处理后本征模态分量IMF中信号难以有效提取的问题,提出改进EEMD-小波阈值的信号处理方法。首先,信号经EEMD处理后得到若干IMF,利用相关性分析提取IMF中的有效分量;在广义交叉验证准则GCV求解阈值的基础上,利用Memetic算法对阈值进行寻优;最后,经阈值函数处理后,重构信号达到信号提取目的。非线性信号处理仿真和实测结果证明了所提方法的有效性。In order to solve the problem that it is difficult to effectively extract the intrinsic mode function(IMF)after the overall ensemble empirical mode decomposition(EEMD)processing,an improved signal processing method is proposed based on the EEMD wavelet-threshold algorithm.Firstly,the signals are processed by the EEMD to obtain several IMFs,and the effective components in the IMFs are extracted through the correlation analysis.Secondly,based on the generalized cross validation(GCV),the Memetic algorithm is used to optimize the threshold.Finally,after the threshold function is processed,the signals are reconstructed to achieve signal extraction.The nonlinear signal processing simulation and actual results indicate that the proposed method is effective.

关 键 词：EEMD 小波阈值 MEMETIC算法 非线性 

分 类 号：TN957.51[电子电信—信号与信息处理]