基于最小二乘支持向量机的电力变压器故障诊断

油中溶解气体分析是变压器绝缘故障诊断的重要方法。为了提高分类的准确度和可靠性,应用最小二乘支持向量机理论建立了变压器的分类模型。该模型以变压器油中5种主要特征气体作为输入量,以7种变压器状态作为输出量,选用了径向基核,使用了一对一的多分类算法,充分发挥了支持向量机具有较高泛化能力的优势。通过大量的实例分析,并将诊断结果与IEC三比值法、改良三比值法和BP神经网络的诊断结果相比较,表明基于径向基核的最小二乘支持向量机在变压器故障诊断中具有更高的准确率。

Power transformer fault diagnosis based on least square support vector machine

DGA(Dissolved Gas Analysis) is essential to the fault diagnosis of power transformer.A model using LS-SVM(Least Square Support Vector Machine) is built to improve the accuracy and reliability of classification.It takes five characteristic gases dissolved in transformer oil as its inputs and seven transformer states as its outputs,selects the radial kernel,applies the one-against-one algorithm,and fully uses the superiority of SVM in processing finite samples.A mass of fault samples are analyzed and results are compared with those obtained by the methods of IEC three-ration,improved three-ration and BPNN,which shows that the LS-SVM algorithm with radial kernel has higher precision.