Author:
Shi Xiaowei,Qu Jinxiu,Guo Jinzhu,Huang Jiaqi,Bai Yumei,Cao Wei
Abstract
Abstract
The viscoelastic sandwich structure exhibits highly nonlinear dynamic characteristics, and minimal changes in signal caused by aging state alterations, thus posing a challenge for extracting its aging characteristics. A method is proposed for extracting aging features from viscoelastic sandwich structures using variational mode decomposition (VMD) and permutation entropy (PE). First, the original dynamic response signal is decomposed by VMD to acquire several intrinsic mode functions (IMF). Then the permutation entropy characteristics of each IMF are calculated. The research results show that there are obvious differences in the distribution of permutation entropy corresponding to each IMF component, which indicates that the permutation entropy feature extracted based on the variational modal decomposition is more susceptive to the change of the aging state of the structure, thus verifying the availability of this method for feature extraction.
Subject
Computer Science Applications,History,Education
Reference8 articles.
1. Aging state detection of viscoelastic sandwich structure based on ELMD and sensitive IA spectrum entropy;Qu;IEEE Access,2019
2. Variational mode decomposition;Dragomiretskiy;IEEE Trans. Signal Process,2014
3. Parameter optimized variational mode decomposition method with application to incipient fault diagnosis of rolling bearing;Tang;J. Xi’An Jiaotong Univ.,2015
4. Application of variational mode decomposition and improved adaptive resonance technology in bearing fault feature extraction;Li;J. Vib. Eng,2018
5. Permutation entropy: a natural complexity measure for time series;Bandt;Phys. Rev. Lett,2002