摘要
针对电力系统运行中对于电力变压器进行高效、准确的故障诊断需要,文中提出了一种基于卷积神经网络的电力变压器故障诊断方法。利用电力变压器油中溶解气体分析法得到特征气体并重新对其进行二进制编码,对编码后的数据进行预处理从而得到特征向量。以特征向量为基础,构建相应的卷积神经网络模型,实现对电力变压的故障诊断。实验结果表明,相较于其他传统机器学习算法,文中所提出的卷积神经网络模型在电力变压器故障诊断时的诊断准确性与诊断效率均有显著优势,能够有效保障电力系统运行的可靠性。
In order to meet the need of high efficiency and accuracy of power transformer fault diagnosis in power system operation,this paper proposes a method of power transformer fault diagnosis based on convolutional neural network.The method of dissolved gas analysis in power transformer oil is used to get the characteristic gas and re-binary code it.The data after coding is pre-processed to get the characteristic vector.Based on the characteristic vector,the corresponding convolution neural network model is constructed to realize the fault diagnosis of power transformer.The experimental results show that compared with other traditional machine learning algorithms,the convolution neural network model proposed in this paper has significant advantages in the accuracy and efficiency of power transformer fault diagnosis,and can effectively guarantee the reliability of power system operation.
作者
夏洪刚
郭红兵
肖金超
XIA Hong-gang;GUO Hong-bing;XIAO Jin-chao(Inner Mongolia Electric Power(Group)Co.,Ltd.,Hohhot 010020,China;Inner Mongolia Electric Power Research Institute,Hohhot 010020,China;Nanjing Qizheng Information Technology Co.,Ltd.,Nanjing 210000,China)
出处
《电子设计工程》
2020年第13期189-193,共5页
Electronic Design Engineering
基金
内蒙古电力科学研究院科技储备项目(2019022)。
关键词
卷积神经网络
电力变压器
故障诊断
特征向量
特征气体
机器学习
convolutional neural network
power transformer
fault diagnosis
characteristic vector
characteristic gas
machine learning
