Open AccessResidual Stress Recognition Method for Welded Structures based on An Improved Multiple Differential Empirical Mode Decomposition
Author(s) -
Xiaohan Liu,
Guangfeng Shi,
Weina Liu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1885/4/042037
Subject(s) - hilbert–huang transform , residual stress , residual , welding , vibration , structural engineering , entropy (arrow of time) , mode (computer interface) , algorithm , mathematics , materials science , energy (signal processing) , computer science , engineering , acoustics , statistics , physics , composite material , thermodynamics , operating system
To overcome the difficulty in accurately identifying the residual stress of complex large welded parts, a multiple differential empirical mode decomposition (MDEMD) method based on grey mean GM(1,1) prediction and mirror symmetry (MS) extension combined with energy entropy is proposed to identify the residual stress state of components. Firstly, the vibration signal of the welded steel plate is collected by the hammering method and preprocessed by MD algorithm. Secondly, GM(1,1) is used to predict the extreme points at the endpoints and mirror symmetry continuation is used to obtain the intrinsic mode function (IMF). Then, the energy value of IMF component is calculated and normalized by energy entropy, and the characteristic parameters are extracted to realize the identification of the residual stress state. The experimental results show that this method can effectively identify the residual stress state of welded components, and the recognition rate reaches 98.7%.