Study of fault diagnosis method based on ensemble-multi-SVM classifiers

Author

Lv Feng ; Li Xiang ; Sun Hao ; Du Hailian ; Rong Wenjie

Author_Institution

Electron. Dept., Hebei Normal Univ., Shijiazhuang, China

fYear

2014

fDate

28-30 July 2014

Firstpage

3272

Lastpage

3276

Abstract

In order to improve the system accuracy of fault diagnosis, this paper proposes the integrated fault diagnosis method based on multi-SVM classifiers. MultiBoost integrated learning method using the AdaBoost algorithm and Wagging algorithm composed of multiple integrated with a combination of base classifiers to improve the classification accuracy of the system. The simulation results show that the method used in network fault diagnosis system of classification module design, making fault diagnosis accuracy has been significantly improved.

Keywords

fault diagnosis; learning (artificial intelligence); pattern classification; support vector machines; AdaBoost algorithm; MultiBoost integrated learning method; Wagging algorithm; classification accuracy; classification module design; ensemble multiSVM classifiers; integrated fault diagnosis method; network fault diagnosis system; support vector machines; Accuracy; Equations; Fault diagnosis; Kernel; Logistics; Support vector machines; Training; Classification; Ensemble Learning; Fault Diagnosis; Support Vector Machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Control Conference (CCC), 2014 33rd Chinese

Conference_Location

Nanjing

Type

conf

DOI

10.1109/ChiCC.2014.6895479

Filename

6895479