摘要
提高变压器故障诊断率及诊断速度是故障诊断研究领域的热点,针对变压器的故障诊断提出了应用KPCA优化神经网络的变压器故障诊断的方法。利用核主元分析方法降维,简化网络结构,使用降维样本来训练网络并识别变压器的故障,并将该方法和传统BP网络进行仿真比较。仿真结果中该方法训练代数降低,识别率提高,速度得到大幅提升。
For the fault diagnosis of transformers,a method of transformer fault diagnosis using KPCA optimized neural network is proposed in this paper.The kernel principal component analysis method was used to reduce the dimension,simplify the network structure,and used the dimensionality reduction sample to train the network and identify the fault of the transformer.The method is compared with the traditional BP network.
出处
《工业控制计算机》
2018年第10期33-34,36,共3页
Industrial Control Computer
基金
国家自然科学基金资助项目(61203343)
关键词
核主元分析
神经网络
速度
故障识别率
变压器故障诊断
kernel principal component analysis
neural network
speed
fault recognition rate
transformer fault diagnosis
