摘要
在决策树基础上 ,建立起适用于变压器故障诊断的组合神经网络模型 ,提出了一种基于多元统计分析技术的训练样本及输入矢量选择方法 ,并给出了网络验证与仿真结果。
A combinatorial neural network model used for transformer fault diagnosis is set up based on the decision tree. A selection method of training samples and input vectors based on the multivariate statistical analysis is pointed ont. The network verification and artificial results are given.
出处
《变压器》
北大核心
2000年第4期32-35,共4页
Transformer
关键词
电力变压器
故障诊断
多元统计分析
神经网络
Power transformer
Combinatorial neural network
Gas analysis
Fault diagnosis
Multivariate statistical analysis
