摘要
依托欧氏距离和支持向量机理论,提出了基于欧氏距离的二叉树支持向量机变压器故障诊断方法,建立了基于欧氏距离的二叉树支持向量机故障诊断模型,最后进行了仿真。结果表明,基于欧氏距离的二叉树支持向量机的变压器故障分类模型不但具有较高的分类准确率,而且能够有效的减小基于二叉树支持向量机故障诊断时误差累计现象的发生。
By using the concept of the Euclidean Distance and the theory of the Support Vector Machine(SVM),this paper proposes a method of binary-tree SVM transformer fault diagnosis based on Euclidean-Distance,and sets up a fault diagnosis model on account of Euclidean distance.Then by the experiment of the model simulation,its results show that the present method not only has a high classification accuracy rate,but also greatly reduce the occurrence of the error accumulation which often occurs in the traditional method for binary-tree SVM transformer fault diagnosis.
出处
《电测与仪表》
北大核心
2013年第6期1-3,19,共4页
Electrical Measurement & Instrumentation
基金
宁夏自治区科学基金资助项目(NZ12140)
委托基金资助项目(411-0318)
关键词
变压器
欧氏距离
支持向量机
故障诊断
transformer
euclidean distance
Support Vector Machine(SVM)
fault diagnosis