基于人工智能的配网变压器故障诊断方法  

Fault Diagnosis Method Based on the Distribution Network Transformer of Artificial Intellence

作　　者：谢孝南 XIE Xiao-nan(Yixin Electric Power Engineering Co.Ltd.of Guangdong,Foshan 528200,China)

机构地区：[1]广东怡信电力工程有限公司,广东佛山528200

出　　处：《电气开关》2022年第4期28-30,35,共4页Electric Switchgear

摘　　要：配网变压器具有控制电力设备的作用,但是在实际应用的过程中时常出现故障与异常,因此,对基于人工智能的配网变压器故障诊断方法进行设计。进行特征故障诊断参量选取,设计智能合环转供故障诊断模型,采用代价矩阵完成对智能变压器的故障诊断。最终测试结果表明:对比于传统AdaBoost敏感故障诊断测试组和传统并行化故障诊断测试组,本文所设计的人工智能故障诊断测试组最终得出的最大绝对差值相对较小,表明此方法的应用故障诊断误差得到了极大地控制,效果更佳,具有实际的应用价值。A disribution network transformer has an effect to control power equipment,but fandts and abnormity often happen in the course of actual application.The fault diagnosis method based on the distribution transformer of artificial intellence is designed.Carry out characteristic fault diagnosis parameter selection,design intelligent combination to provide the fault disagnosis model and adopt price matrix to finish the fault diagnosis to intelligent transformer.The final test result shows that that the errors are greatly controlled compared with traditional AdaBoost sensitive fault diagnosis test group and traditional parallel fault diaghosis test group.The artificial intelligence faut diagnosis test group designed in the paper finally obtain maximum absolute difference value smaller,which shows the use of the method fault diagnosis error is greatly controlled.The effect is better and the method has actual application value.

关 键 词：人工智能 配网结构 变压器 故障诊断 诊断方法 智能控制 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]