A fuzzy based semi-supervised method for fault diagnosis and performance evaluation

Author

Yixiang Huang ; Liang Gong ; Shuangyuan Wang ; Lin Li

Author_Institution

Sch. of Mech. Eng., Shanghai Jiao Tong Univ., Shanghai, China

fYear

2014

fDate

8-11 July 2014

Firstpage

1647

Lastpage

1651

Abstract

How to automatically deal with the unknown classes or status of a machine is a practical problem in many real-world applications. One of the key solutions is to enable the intelligent models with learning ability. Neither supervised nor unsupervised methods can well handle it. In this paper, we proposed a fuzzy based semi-supervised method to not only make the best of the known knowledge but also category the unknown status in a reasonable way. A roller bearing test validates the proposed method for the purpose of both diagnosis and performance evaluation.

Keywords

fault diagnosis; fuzzy set theory; mechanical engineering computing; mechanical testing; performance evaluation; reliability; rolling bearings; unsupervised learning; fault diagnosis; fuzzy based semisupervised method; intelligent models; learning ability; performance evaluation; roller bearing test; unsupervised methods; Educational institutions; Fault diagnosis; Mechanical systems; Performance evaluation; Signal processing; Training; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Advanced Intelligent Mechatronics (AIM), 2014 IEEE/ASME International Conference on

Conference_Location

Besacon

Type

conf

DOI

10.1109/AIM.2014.6878320

Filename

6878320