经验模态分解、数值仿真与形变信号滤波试验

Empirical mode decomposition,data emulation and test of deformation signal filtering

首先介绍了经验模态分解的方法与原理,给出了由经验模态分解产生的固有模态函数重构特定滤波器的算法,然后通过对已知数据的仿真检验,认为该方法在分解与重构滤波器方面具有较高的精度,最后通过对形变数据中几类典型干扰信号的初步试验与对比分析,可看出获得了满意的处理效果。综合分析认为,经验模态分解及其组合滤波器在干扰消除或某些频率成分滤波与研究中具有重要意义,可作为形变数据预处理中一种行之有效的数据分解方法。

Firstly, the theory of empirical mode decomposition （EMD） is introduced, and the reconstruction and filtering algorithm of intrinsic mode functions （IMF） from EMD is given in this paper. Secondly, effective experiment is done by using the emulation data. Lastly some kinds of typical interfere signals are decomposed and filtered by using EMD technique in deformation observation. By means of contrasting and analyzing, we conclude that the result is fairly effective. It has shown that EMD may be a new effective processing method in analyses of deformation data.