基于经验模态分解的Hilbert-Huang变换在分解过程中产生端点效应。使得信号两端点附近出现失真问题,提出对端点效应进行抑制的多种方法.对全局统计平均法、平行线段延拓极值法、多项式拟合法、镜像延拓法、神经网络延拓法等抑制方法进行对比研究.仿真和实验结果表明,端点效应得到有效抑制,Hilbert-Huang变换的分析性能得到改善.
Ending effect is produced in the decomposition process of empirical mode decomposition (EMD) based Hilbert-Huang transformation. It leads to distortion near the two end points of the signal processed. Several methods of restraint aimed at ending effect are proposed. Comparison has been made among the global average statistic method, the parallel line extending method, the polynomial curvefitting method, the mirror extending method and the neural networks extending method. As a result of simulation and experiment, it is shown that ending effect can be effectively restrained.