摘要
考虑到构建二叉树支持向量机时样本的分布情况对分类器推广能力具有较大影响,提出一种改进的二叉树支持向量机层次结构构建方法.以类间样本距离和带权值的类内样本距离与其标准差的比值作为类的分类度.将类间距离大且类内样本平均分布广的类最先分离.利用标准数据集,通过与不同多类分类算法比较,验证了改进的二叉树支持向量机的优越性.对双转子涡喷发动机气路部件进行应用改进的算法进行故障诊断,得到了较好的故障识别率.
Considering of the sample range was proposed to rationally array classifiers of a support vector machine with binary tree architecture. An improved binary tree SVM hierarchy construction method was presented. A separability measure with weights was constructed by the ratio of the average distance of samples in one class with the standard deviation of the samples from the same class and the average distance of samples between different classes. So the class which has bigger distance from other classes and wider average sample distribution within itself was first separated. Compared with recognition accuracy of standard data sets for differentmulti-class algorithm, the superiority of improved binary tree SVM is verified. A simulation diagnosis experiment for the gas path components of a turbojet engine is conducted to demonstratethe effect of the improved algorithm and the fault classifiers have good accuracy.
作者
王龚
张金春
盖明久
WANG Yan ZHANG Jin-chun GAI Ming-jiu(Postgraduate, Naval Aeronautical ＆ Astronautical University, Yantai 264001, China Naval Aeronautical and Astronautical University, Yantai 264001, China)
出处
《数学的实践与认识》
北大核心
2017年第19期92-98,共7页
Mathematics in Practice and Theory
关键词
二叉树
支持向量机
故障诊断
binary tree
support vector machines
fault diagnosis
