基于神经网络的变压器早期故障诊断安全技术

Transformer's Trouble Diagnosis Safety Technology Based on Neural Network

介绍了油中气体分析法中最常用的判断方法———罗杰斯三比值法。应用BP神经网络方法 ,建立变压器故障检测的BP网络模型 ,经过训练 ,对变压器故障进行早期诊断 ,运行结果与实际相符。

This paper introduces Royce three ration method that is getting wide notice in analytical method of gas behind oil. By using BP network, the transformer's trouble diagnosis model is formed. In the end , we train the BP network to diagnose transformer's trouble according to three ration method. The result shows agreement with practice.