To solve the problems of insufficient fault-samples and diagnosis-knowledge, and according to the merit of Support Vector Machines (SVM) that can be trained with small-sample, a SVM based unsupervised clustering model is presented. By modifying the decision-function of One-Class Support Vector Machine (1-SVM), which has the ability to find outliers from a dataset without any class of information but rarely is applied to pattern-recognition for its algorithm limits, a Decision-Improved 1-SVM (1-DISVM) is formed. Based on it, multi-pattern training and classing method is designed, then an unsupervised clustering model is constructed. The simulation and diagnostic experiment results of a helicopter��s gearbox show that this clustering model can not only recognize the unknown fault patterns adaptively and precisely, but also learn the nature of the input-patterns from small samples and diagnose the faults successfully.
LIU Xin-min;LIU Guan-jun;QIU Jing;HU Niao-qing. Unsupervised 1-DISVM Based Clustering Model for Fault Diagnosis of Helicopter Gearbox[J]. Acta Aeronautica et Astronautica Sinica, 2006, 27(3): 453-458.
 ʷ����,���»�,��Ρ��,��. ��ߴ����ӻ��ϵͳ������������ļ����·���[J]. ��е����ѧ��,2003, 39(9):1-10. Shi T L, Chen Y H, Li W H, et al. Some measures to improve the quality of fault diagnosis for large-scale complex electromechanical systems [J]. Chinese Journal of Mechanical Engineering, 2003, 39(9):1-10.(in Chinese)  Vapnik V N.ͳ��ѧϰ���۵ı���[M]. ��ѧ����.����: �廪��ѧ������, 2000. Vapnik V N. The nature of statistical learning theory��[M]. Translated by Zhang X G. Beijing: Qinghua University Press, 2000. (in Chinese)  Batur C,Zhou L,Chan C C. Support vector machines for fault detection��. Proceedings of the 41st IEEE Conference on Decision and Control��.Las Vegas,Nevada USA, 2002. 1355-1356.  Salat R,Osowski S. Accurate fault location in the power transmission line using support vector machine approach��[J]. IEEE Transactions on Power Systems,2004,19(2):979-986.  Sch��lkopf B,Platt J,Shawe-Taylor J, et al. Estimating the support of a high-dimensional distribution��[J]. Neural Computation,2001,13:1443-1471.  Tran Q A,Zhang Q L,Li X. Evolving training model method for one-class SVM��. In:IEEE International Conference on Systems, Man and Cybernetics��.Washington D C, 2003. 2388-2393.  ����Т. ʱ�����з����� [M]. �人:�����?��ѧ������,1991. Li X X. Analysis base of time serial [M].Wuhan: Science and Industry University of Central China Press,1991.(in Chinese)