An adaptive variational mode decomposition based on sailfish optimization algorithm and Gini index for fault identification in rolling bearings

•Sailfish algorithm (SFO) is used to optimize VMD parameters (k) and (α).•Gini index is used in SFO as a fitness function as it is sensitive to impulses.•Determination of the relevant mode was carried out based on Gini index.•The proposed method can optimize the VMD parameters for high decomposition...

Full description

Saved in:
Bibliographic Details
Published in:Measurement : journal of the International Measurement Confederation 2021-03, Vol.173, p.108514, Article 108514
Main Authors: Nassef, M.G.A., Hussein, Taha M., Mokhiamar, Ossama
Format: Article
Language:English
Subjects:
Algorithms
Bearings
Decomposition
Fault detection
Fault diagnosis
Feature extraction
Gini index
Noise reduction
Optimization
Parameters
Roller bearings
Rolling element bearing
Sailfish algorithm
Signal processing
Variational mode decomposition
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Sailfish algorithm (SFO) is used to optimize VMD parameters (k) and (α).•Gini index is used in SFO as a fitness function as it is sensitive to impulses.•Determination of the relevant mode was carried out based on Gini index.•The proposed method can optimize the VMD parameters for high decomposition quality. Variational mode decomposition (VMD) method is a recently employed signal processing technique for fault identification from the vibration signal of rolling bearings. However, the selection of VMD parameters namely the number of modes (k) and the quadratic penalty factor (α) still represents a challenge to obtain proper decomposition modes with the most relevant and denoised fault information. This paper presents a framework using sailfish optimization (SFO) algorithm and Gini index (GI) as a criterion to adaptively select the optimum VMD parameters for each fault signal. The proposed algorithm is tested using three experimental signals of faulty bearings, and the most appropriate mode containing fault information is automatically extracted based on maximum GI values. The obtained results indicate high efficiency of the proposed method in extracting fault feature and in exclusion of noise effect as compared to conventional fixed-parameter VMD, local mean decomposition (LMD), and ensemble empirical mode decomposition (EEMD).
ISSN:0263-2241
1873-412X
DOI:10.1016/j.measurement.2020.108514