摘要
变压器是电力系统的一个重要设备。为了确保变压器安全运行,杜绝事故的发生,对变压器故障进行诊断就显得尤为重要。笔者提出了一种变压器在线实时故障诊断系统,设计了硬件电路,利用智能型气体传感器和物联网技术建立了数据传输网络,实现对变压器内部参数的智能采集;阐述了系统建模的理论和方法,利用贝叶斯、KNN和决策树等模型对变压器故障进行分类、预测,并比较三种模型的性能。结果表明:KNN与决策树模型对变压器故障分类、预测的正确率达到100%;贝叶斯模型分类、预测的正确率比较低,仅有84%。变压器在线实时故障诊断系统能够安全、稳定地运行,在变压器故障诊断中应用KNN和决策树模型是可行的,可供变压器在线监测和故障诊断参考。
Transformer is an important equipment of power system.In order to ensure the safe operation of transformer and prevent the occurrence of accidents,it is very important to diagnose the transformer fault.In this paper,an online real-time fault diagnosis system for transformer is proposed.The hardware circuit is designed,and the data transmission network is established by using intelligent gas sensors and internet of things technology to realize the intelligent acquisition of internal parameters of the transformer.The theory and method of system modeling are expounded.Bayesian,KNN and decision tree models are used to classify and predict transformer faults,and the performance of the three models is compared.The results show that the accuracy of classification and prediction of transformer faults by KNN and decision tree models reaches 100%;The accuracy of classification and prediction using Bayesian model is relatively low,only 84%.The transformer online real-time fault diagnosis system can operate safely and stably.It is feasible to apply KNN and decision tree models in transformer fault diagnosis,which can provide reference for transformer online monitoring and fault diagnosis.
作者
刘裕舸
LIU Yuge(Liuzhou Railway Vocational Technical College,Liuzhou,Guangxi,545616)
出处
《红水河》
2021年第2期90-95,共6页
Hongshui River
关键词
变压器
在线实时系统
故障诊断
贝叶斯
K近邻
决策树模型
transformer
online real-time system
fault diagnosis
Bayesian
K-nearest neighbor
decision tree model
