Morphological Undecimated Wavelet Decomposition Fusion Algorithm and Its Application on Fault Feature Extraction of Hydraulic Pump |
| |
Authors: | Sun Jian Li Hongru Wang Weiguo Ye Peng |
| |
Institution: | Mechanical Engineering College, Shijiazhuang 050003, P.R.China |
| |
Abstract: | Since vibration signals of hydraulic pump are mostly nonlinear and traditional fusion algorithm cannot satisfyingly process them,a morphological undecimated wavelet decomposition fusion(MUWDF)algorithm is proposed.Firstly,under the framework of morphological undecimated wavelet decomposition(MUWD),multi-channel signals are decomposed.Approximate signals of all decomposition layers are selected by feature energy factor and fused according to the presented fusion rules.Furthermore,specific method for optimal selection of MUWDF parameters is presented to avoid subjective influences.Finally,the proposed algorithm is verified by simulation signals and pump vibration signals. |
| |
Keywords: | morphological undecimated wavelet decomposition (MUWD) information fusion feature extraction hydraulic pump |
本文献已被 CNKI 等数据库收录! |
|