基于SMOTE-SSA-CNN的煤矿用变压器DGA故障诊断方法  

DGA Fault Diagnosis Method for Coal Mine Transformer Based on

作　　者：张化昭 朱建武 邱日强 彭明聪 林江 Zhang Huazhao;Zhu Jianwu;Qiu Riqiang;Peng Mingcong;Lin Jiang(State Grid Nanchang Power Supply Company,Nanchang 330200,China;Nanjing Deruan Information Technology Development Co.,Ltd.,Nanjing 210000,China)

机构地区：[1]国网南昌供电公司,南昌330200 [2]南京德软信息科技发展有限公司,南京210000

出　　处：《煤矿机械》2024年第12期172-176,共5页Coal Mine Machinery

摘　　要：为了提高基于油中溶解气体分析(DGA)的煤矿用变压器故障识别精确性,提出了一种基于SMOTE-SSA-CNN的煤矿用变压器DGA故障诊断模型。首先,以煤矿用变压器油中溶解气体为基础数据,采用合成少数类样本过采样(SMOTE)算法对原始数据集进行样本扩充,解决原始数据集中正负样本严重失衡的问题;然后引入麻雀搜索算法(SSA)对卷积神经网络(CNN)的卷积核大小与数量、全连接层神经元数量、学习率等超参数进行优化,提高模型故障诊断结果的准确率;最后,通过算例分析对建立的SMOTE-SSA-CNN模型性能进行评估,验证了所提方法对煤矿用变压器故障诊断的有效性,且与传统故障诊断方法相比,所提方法的收敛性较好,精度较高。In order to improve the accuracy of coal mine transformer fault identification based on dissolved gas analysis(DGA),proposed a fault diagnosis model based on SMOTE-SSA-CNN for coal mine transformer.Firstly,based on the dissolved gases in coal mine transformer oil,the synthetic minority sample oversampling(SMOTE)algorithm was used to expand the original dataset and solve the problem of severe imbalance between positive and negative samples in the original dataset.Then,the sparrow search algorithm(SSA)was introduced to optimize the hyperparameters of convolutional neural networks(CNN),such as the size and number of convolution kernels,the number of fully connected layer neurons and the learning rate,in order to improve the accuracy of the model fault diagnosis results.Finally,the performance of the established SMOTE-SSA-CNN model was evaluated through numerical examples,and verified the effectiveness of the proposed method for fault diagnosis of coal mine transformer.Compared with traditional fault diagnosis methods,the proposed method has better convergence and higher accuracy.

关 键 词：煤矿用变压器 DGA SMOTE SSA CNN 

分 类 号：TD611[矿业工程—矿山机电]