基于模糊支持向量机的多级二叉树分类器的水轮机调速系统故障诊断张国云.pdfVIP

25 8 Vol.25 No.8 Apr. 2005 2005 4 Proceedings of the CSEE ©2005 Chin.Soc.for Elec.Eng. 0258-8013 (2005) 08-0100-05 TM41 A 470 20 ⋅ 1,2 1 (1. 410082 2. 414006) FUZZY SVM-BASED MULTILEVEL BINARY TREE CLASSIFIER FOR FAULT DIAGNOSIS OF HYDROTURBINE SPEED REGULATING SYSTEM ZHANG Guo-yun1,2 , ZHANG Jing1 (1. College of Electrical Information Engineering, Hunan University, Changsha 410082, Hunan Province, China; 2.Dept. of Electronics Information, Hunan Institute of Science Technology, Yueyang 414006, Hunan Province, China) ABSTRACT: Based on conventional Support Vector k k-1 SVM Machine(C-SVM) and through integrating fuzzy clustering technique and SVM algorithm, a multilevel binary tree classifier which is suitable for fault diagnosis task is presented and applied to fault diagnosis task of hydroturbine speed regulating system for the first time. Firstly, this method computes the clustering centers of each class by using fuzzy 1 clustering technique and all clustering centers are divided into two successively. So, a binary tree is constructed. Then, it reconstructs SVM sub-classifier according to the clustering centers and the samples which belong to those clustering BP centers in each node of binary tree. The experimental results [1]BP show that, in k-class fault diagnosis task, only k-1 SVM sub-classifiers need