Research on Classification Optimization Algorithm of Rotor to Stator Rub-impact Fault Using FGSO-SVM

Rotor to stator rub-impact fault (RSRIF) is a common fault in rotating machinery. In this paper, a FGSO-SVM algorithm is proposed for the classification and recognition of RSRIF. The algorithm, combining glowworm swarm optimization algorithm (GSO) and shuffled frog-leaping algorithm (SFLA), optimize...

Full description

Saved in:
Bibliographic Details
Published in2020 International Symposium on Computer Engineering and Intelligent Communications (ISCEIC) pp. 103 - 106
Main Authors You, Lei, Wang, Jin, Han, Qiyi, Chen, Jing, Liang, Ying
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.08.2020
Subjects
classificaiton
FGSO-SVM
Machinery
Optimization
Particle swarm optimization
rotating machinery
Rotors
rub-impact fault
Stators
Zirconium
Online AccessGet full text
DOI10.1109/ISCEIC51027.2020.00029

Cover

More Information
Summary:Rotor to stator rub-impact fault (RSRIF) is a common fault in rotating machinery. In this paper, a FGSO-SVM algorithm is proposed for the classification and recognition of RSRIF. The algorithm, combining glowworm swarm optimization algorithm (GSO) and shuffled frog-leaping algorithm (SFLA), optimizes the support vector machine (SVM) parameters. Compared with traditional algorithms, FGSO-SVM improves the recognition rate of fault classification and overcomes the deficiencies of being easily trapped into local optimum and slow late convergence rate. Three algorithms herein, namely BPNN, AFSA-SVM and FGSO-SVM, are used for a comparative study under normal condition and RSRIF. The experimental results show that FGSO-SVM has the highest classification and recognition rate of RSRIF.
DOI:10.1109/ISCEIC51027.2020.00029