分离EMD中混叠模态的新方法  

A new method for separating mixed modes in empirical mode decomposition

作　　者：高云超[1] 桑恩方[1] 许继友 

机构地区：[1]哈尔滨工程大学水声工程学院,黑龙江哈尔滨150001 [2]船舶系统工程部七分部,北京100086

出　　处：《哈尔滨工程大学学报》2008年第9期963-966,共4页Journal of Harbin Engineering University

摘　　要：针对经验模态分解(empirical mode decomposition,EMD)过程中可能出现的模态混叠问题,提出一种新的处理方法.该方法利用了差分运算、累计求和,同时结合EMD来实现混叠模态的分离.该文对数值仿真信号和LOD(length-of-day)数据进行了仿真分析,比较了原始EMD和所提算法的分析结果.仿真结果表明,该算法可有效分离混叠模态,得到尺度成分清晰的本征模态函数.A new method has been proposed to solve the problem of mode mixing in empirical mode decomposition （EMD）. The method combines the difference operation, the cumulative sum, and EMD to perform the separation of mixed modes. A computer simulation was made on numerical simulation signals and LOD （length of day） data. A comparison between the results from EMD algorithm and the proposed algorithm was executed. The results showed that the proposed method can separate the mixed mode successfully and a set of Intrinsic Mode Functions （IMF） with clear scale range can be obtained.

关 键 词：信号分析 时间序列分析 经验模态分解 模态混叠 

分 类 号：TN911.72[电子电信—通信与信息系统]