A Novel Fault Class Detection Method using Multi-class Support Vector Machines with Case-Based Reasoning Combination

다국어 초록 (Multilingual Abstract)

his paper explores a novel fault class detection method using a hybrid multi-class support vector machines and case-based reasoning approach. The multi-class classification method usually has a main shortcoming that the binary classifiers used are obtained by training on different binary classification problems, and thus it is unclear whether their real-valued outputs are on comparable scales. In this paper, we try to use additional information, relative outputs of the machines, for final decision. We propose case-based reasoning with reject option to use the information. The experiments on real multi-class rolling bearing fault data of classification process shows that the proposed method achieves a better performance than the previous multi-class support vector machines.

참고문헌 (Reference)

1 Tax, M. J., "Support vector data description" 54 : 45-66, 2004

2 Mayoraz, E., "Support Vector Machines for Multi-class Classification" IDIAP 1998

3 Vapnik, V.N., "Statistical learning theory" John Wiley & Sons 1998

4 Markou, M., "Novelty detection: a review-part 2: neural network based approaches" 83 (83): 2499-2521, 2003

5 Markou, M., "Novelty detection: a review-part 1: statistical approaches" 83 (83): 2481-2497, 2003

6 Scholkopf, B., "Learning with Kernels" MIT Press 2002

7 Thomson, W. T., "Current signature analysis to detect induction motor faults" 7 : 26-34, 2001

8 Lou, X., "Bearing fault diagnosis based on wavelet transform and fuzzy inference" 18 (18): 373-390, 2004

9 Thollon, F., "Asynchronous motor cage fault detection through electromagnetic torque measurement" 3 : 375-378, 1993

10 Stone, G. C., "Application of partial discharge testing to motor and generator stator winding maintenance" 32 : 459-464, 1996

11 Lee, Y. S., "An acoustic diagnostic technique for use with electric machine insulation" 9 : 1186-1193, 1994