Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation

Compound faults easily happen in rolling bearing due to the complex working environment. Diagnosing compound faults accurately is a thorny problem, which can ensure the normal operation of mechanical structure. To tackle this problem, this paper proposes a novel method called double-domain reweighte...

Full description

Saved in:
Bibliographic Details
Main Authors: Jing Meng, Jiawen Xu, Chang Liu, Chao Chen, Lili Liu
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10802894/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846109889330413568
author Jing Meng
Jiawen Xu
Chang Liu
Chao Chen
Lili Liu
author_facet Jing Meng
Jiawen Xu
Chang Liu
Chao Chen
Lili Liu
author_sort Jing Meng
collection DOAJ
description Compound faults easily happen in rolling bearing due to the complex working environment. Diagnosing compound faults accurately is a thorny problem, which can ensure the normal operation of mechanical structure. To tackle this problem, this paper proposes a novel method called double-domain reweighted adaptive sparse representation. The proposed method reweights the weight through the fault information from the wavelet domain and time domain. In the wavelet domain, periodic clustering similarity is proposed to measure fault features by similarity calculation. Then, wavelet domain information is transformed into time domain. In time domain, the Hadamard product of the sub-band signals is calculated to obtain fault features. The fault features in time domain are further transferred into wavelet domain, which is used for reweighting the weight. The proposed model’s optimal parameters are determined through the hippopotamus optimization algorithm, with the proposed frequency correlated kurtosis as the objective function. Frequency correlated kurtosis measures the signal’s frequency domain periodicity. The proposed method is verified by simulation, experimental signals, and comparison experiments, whose results show the effectiveness and superiority of the proposed method.
format Article
id doaj-art-e3af63e426ee4115a3e8ef7fb748d8fc
institution Kabale University
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-e3af63e426ee4115a3e8ef7fb748d8fc2024-12-25T00:01:32ZengIEEEIEEE Access2169-35362024-01-011219329919331210.1109/ACCESS.2024.351783310802894Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse RepresentationJing Meng0https://orcid.org/0000-0001-7973-1225Jiawen Xu1https://orcid.org/0000-0002-5398-0394Chang Liu2Chao Chen3https://orcid.org/0000-0002-1410-5466Lili Liu4School of Mechanical and Electrical Engineering, Xuzhou University of Technology, Xuzhou, ChinaSchool of Instrument Science and Engineering, Southeast University, Nanjing, ChinaSchool of Mechanical and Electrical Engineering, Xuzhou University of Technology, Xuzhou, ChinaSchool of Mechanical Engineering, Jiangsu University, Zhenjiang, ChinaSchool of Mechanical and Electrical Engineering, Xuzhou University of Technology, Xuzhou, ChinaCompound faults easily happen in rolling bearing due to the complex working environment. Diagnosing compound faults accurately is a thorny problem, which can ensure the normal operation of mechanical structure. To tackle this problem, this paper proposes a novel method called double-domain reweighted adaptive sparse representation. The proposed method reweights the weight through the fault information from the wavelet domain and time domain. In the wavelet domain, periodic clustering similarity is proposed to measure fault features by similarity calculation. Then, wavelet domain information is transformed into time domain. In time domain, the Hadamard product of the sub-band signals is calculated to obtain fault features. The fault features in time domain are further transferred into wavelet domain, which is used for reweighting the weight. The proposed model’s optimal parameters are determined through the hippopotamus optimization algorithm, with the proposed frequency correlated kurtosis as the objective function. Frequency correlated kurtosis measures the signal’s frequency domain periodicity. The proposed method is verified by simulation, experimental signals, and comparison experiments, whose results show the effectiveness and superiority of the proposed method.https://ieeexplore.ieee.org/document/10802894/Compound fault diagnosisdouble-domainadaptive sparse representation
spellingShingle Jing Meng
Jiawen Xu
Chang Liu
Chao Chen
Lili Liu
Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
IEEE Access
Compound fault diagnosis
double-domain
adaptive sparse representation
title Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
title_full Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
title_fullStr Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
title_full_unstemmed Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
title_short Bearing Compound Fault Diagnosis Based on Double-Domain Reweighted Adaptive Sparse Representation
title_sort bearing compound fault diagnosis based on double domain reweighted adaptive sparse representation
topic Compound fault diagnosis
double-domain
adaptive sparse representation
url https://ieeexplore.ieee.org/document/10802894/
work_keys_str_mv AT jingmeng bearingcompoundfaultdiagnosisbasedondoubledomainreweightedadaptivesparserepresentation
AT jiawenxu bearingcompoundfaultdiagnosisbasedondoubledomainreweightedadaptivesparserepresentation
AT changliu bearingcompoundfaultdiagnosisbasedondoubledomainreweightedadaptivesparserepresentation
AT chaochen bearingcompoundfaultdiagnosisbasedondoubledomainreweightedadaptivesparserepresentation
AT lililiu bearingcompoundfaultdiagnosisbasedondoubledomainreweightedadaptivesparserepresentation