A Hybrid Multi-Experts Approach for Mechanical Defects´ Detection and Diagnosis

Author

Sene, Mbaye ; Chebira, Abdennasser ; Madani, Kurosh

Author_Institution

UFR SAT, Gaston Berger Univ., St. Louis

fYear

2008

fDate

26-28 June 2008

Firstpage

59

Lastpage

64

Abstract

Compared with parametric classifiers, several advantages set neural networks as privileged approaches to be used as discriminating classifiers in performing diagnosis tasks. In this paper, we present a hybrid multi-experts neural based architecture for mechanical defects´ detection and diagnosis. This solution is evaluated within vibratory analysis frame using a wavelet transform faults´ detection scheme.

Keywords

fault diagnosis; mechanical engineering computing; vibrations; wavelet transforms; diagnosis tasks; hybrid multiexperts approach; mechanical defects detection; mechanical defects diagnosis; neural networks; vibratory analysis; wavelet transform faults detection scheme; Artificial intelligence; Electrical fault detection; Monitoring; Neural networks; Shape; Signal analysis; Signal processing; Turning; Wavelet analysis; Wavelet transforms; Artificial Intelligence; Fault Detection; Fault Diagnosys; Hybrid system; Mechanical Plants;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Information Systems and Industrial Management Applications, 2008. CISIM '08. 7th

Conference_Location

Ostrava

Print_ISBN

978-0-7695-3184-7

Type

conf

DOI

10.1109/CISIM.2008.57

Filename

4557835