摘要
现有的电力变压器故障检修方法自动化程度较低。为此,引入多故障分类技术,对现有电力变压器故障检修方法进行优化。采用KPCA方法提取电力变压器数据特征,以数据特征为基础,选取核函数,依据最小二乘支持向量机设计多故障分类器,确定电力变压器故障类别与代码,依据多故障分类器以及电力变压器故障类别,搭建电力变压器故障自动检修模型,实现了电力变压器故障的自动检修。结果表明:相较于现有的电力变压器故障自动检修方法,提出的电力变压器故障自动检修方法极大地提升了自动化程度,具备更好的故障检修性能。
In view of low degree of automation of existing trouble shooting approaches for power transformers,a multi-fault classification technology was introduced to optimize currently used methods.Data characteristics of transformers were extracted in the KPCA method,and based on those data characteristics,kernel functions were selected.A multi-fault classifier was designed according to the least squares support vector machine(SVM)to determine the fault type and code of the power transformer.An automatic trouble shooting model was established for the power transformer according to the multi-fault classifier and the fault type of the power transformer,thus realizing automatic trouble shooting of the power transformer.The results showed that,compared with the existing automatic trouble shooting method for the power transformer,the proposed automatic trouble shooting method for the power transformer could greatly raise the degree of automation and have better trouble shooting performance.
作者
贾文皓
Jia Wenhao(State Grid Zhejiang Jinyun County Power Supply Co., Ltd., Jinyun Zhejiang 321400, China)
出处
《电气自动化》
2021年第2期109-111,118,共4页
Electrical Automation