摘要
对RBF神经网络的基本原理进行了介绍,研究利用了油浸式变压器的状态能够根据油中各种气体含量来判断的特性,利用RBF具有局部任意精度逼近的特点,建立了与变压器油中气体含量相关的人工神经网络状态识别模型。该模型作为在线监测的一种有效辅助工具可以准确的识别出变压器的状态。
The paper introduced the basic principle of RBF neural network and researched judgment method to measure the oil im-mersed type transformer condition based on the various gas contents. Using the characteristics of RBF regional arbitrary accuracy ap-proximation, the paper constructed the neural network in relation to the gas content in oil immersed type transformer. The method could recognize the transformer condition and be one efficient tool of online monitoring.
出处
《贵州电力技术》
2013年第4期33-35,72,共4页
Guizhou Electric Power Technology
关键词
RBF神经网络
变压器
状态识别
在线监测
RBF neural network
transformer
condition recognize
online monitor
