•  
  •  
 

Journal of System Simulation

Abstract

Abstract: To improve the accuracy of mechanical fault diagnosis, it’s meaningful to do research on the boundary effect of EMD. In this article, least squares linear method is applied to extend local extremum points at the end of a data series, which judges that whether the maximum value is within the wave fluctuation range of the original signal maximum points, whether the minimum value is within the wave fluctuation range of the original signal minimum points. It compares the extended extreme points and endpoint values, so the best signal endpoint is obtained. By calculating the value of similarity coefficient and decomposition error, the results of simulation analysis and tests of fault diagnosis of rolling bearings show that the developed EMD method can depress the boundary effect compared with parallel extending method, extrema shift method and mirror periodic.

First Page

3377

Last Page

3385

CLC

TK422

Recommended Citation

He Zhenpeng, Zhu Zhiqi, Xie Haichao, Wang Yawen, Li Zongqiang, He Rui, Du Chaoping, Li Jinlan. Restrain Boundary Effect of EMD Based on Least Square Fitting[J]. Journal of System Simulation, 2018, 30(9): 3377-3385.

DOI

10.16182/j.issn1004731x.joss.201809019

Share

COinS