基于模糊支持向量机的变压器故障诊断

Research of transformer fault diagnosis based on fuzzy support vector machines

针对支持向量机对训练样本内的噪声和孤立点比较敏感,影响了支持向量机分类性能的弱点,利用模糊支持向量机的学习方法,构建了变压器故障诊断模型.采取一种基于二叉树的多分类方法,使用模糊C均值聚类算法求取模糊支持向量机的模糊隶属度,采用径向基核函数,并利用遗传算法对模糊支持向量机的参数进行寻优.实验结果表明,基于二叉数的模糊支持向量机模型相比BP神经网络、支持向量机有更高的诊断准确率,基于二叉树模糊支持向量机的变压器故障诊断方法是可行的.

As the traditional support vector machines（SVM） is particularly sensitive to noise and outliers in the training samples, transformer fault diagnosis model is built based on fuzzy support vector machines（FSVM）. It selects binary tree classification algorithm for multi-class classification. The membership value of the FSVM is obtained by fuzzy C-means clustering algorithm. The radial kernel is selected. Parameters of the FSVM model are optimized with genetic algorithm. A mass of fault samples are analyzed and results are compared with those obtained by the methods of BPNN and SVM, which shows that the proposed model is more effective and accurate, which proves the given method of transformer fault diagnosis based on binary tree fuzzy support vector machines is feasible.