基于油中气体分析的变压器故障诊断ReLU-DBN方法  

Dissolved Gas Analysis of Insulating Oil for Power Transformer Fault Diagnosis Based on Re LU-DBN

作　　者：代杰杰 宋辉[1] 杨祎 陈玉峰 盛戈皞[1] 江秀臣[1] 

机构地区：[1]上海交通大学电气工程系,上海市闵行区200240 [2]国网山东省电力公司电力科学研究院,山东省济南市250002

出　　处：《电网技术》2018年第2期658-664,共7页Power System Technology

基　　金：国家自然科学基金项目(51477100);国家863高技术基金项目(2015AA050204);国家电网公司科技项目~~

摘　　要：油中溶解气体分析可为变压器故障诊断提供重要依据。为提高变压器故障诊断精度,研究了基于修正线性单元改进的深度信念网络（rectified linear units deep belief networks,Re LU-DBN）变压器故障诊断方法。通过分析油中溶解气体与故障类型的联系,建立以油色谱特征气体无编码比值为特征参量的Re LU-DBN诊断模型。Re LU-DBN通过多维多层映射提取出故障类型更细致明显的特征区别,通过反向调优达到诊断模型参数最优化。通过识别实验分析了不同特征参量、不同训练集及样本集大小下Re LU-DBN诊断模型效果,研究了放电兼过热复合型故障对诊断模型的影响,并与支持向量机、反向传播神经网络方法做了对比。实验结果表明基于无编码比值的模型诊断效果优于IEC比值、Rogers比值、Dornenburg比值为特征参量的模型,且Re LU-DBN较支持向量机和反向传播神经网络方法相比诊断准确率有较大提高。区分复合型故障的模型诊断效果优于未区分复合型故障的模型。随着样本数据的增多,模型诊断精度得到较大提升。Dissolved gas analysis（DGA） of insulating oil can provide an important basis for transformer fault diagnosis. To improve diagnosis accuracy, this paper studies a transformer fault diagnosis method based on rectified linear units deep belief networks（Re LU-DBN）. By analyzing relationship between the gases dissolved in transformer oil and fault types, non-code ratios of the gases are determined as characteristic parameters of Re LU-DBN model. Re LU-DBN adopts multi-layer and multi-dimension mapping to extract more detailed differences of fault types. In this process, the parameters of diagnosis model are pre-trained, and adjusted with back propagation algorithm with sample labels. Performances of the Re LU-DBN diagnosis model are analyzed with different characteristic parameters, different training dataset and sample dataset. Besides, influence of discharge and overheating multiple-fault on diagnosis model is studied. Diagnosis effect of the model with non-code ratios as characteristic parameters is better than those of the models with IEC ratios, Rogers ratios and Dornenburg ratios. Compared with the results derived from support vector machine（SVM） and back propagation neural network（BPNN）, the proposed Re LU-DBN method significantly improves accuracy for power transformer fault diagnosis. Diagnosis effect of the model considering multiple-fault is better than that of the model without multiple-fault. With increase of sample dataset, the diagnostic accuracy is improved.

关 键 词：变压器 油中气体分析 深度信念网络 无编码比值 故障诊断 

分 类 号：TM85[电气工程—高电压与绝缘技术]