RVM和ANFIS用于变压器故障诊断及状态评估  

Fault Diagnosis and Status Evaluation for Transformers Using RVM and ANFIS

作　　者：范竞敏 汪沨[1] 孙秋芹[1] 蒋勤稷 欧明辉 FAN Jingmin;WANF Feng;SUN Qiuqin;JIANG Qinji;OU Minghui(School of Electrical and Information Engineering,Hunan University,Changsha 410082,China)

机构地区：[1]湖南大学电气与信息工程学院,长沙410082

出　　处：《电力系统及其自动化学报》2018年第3期35-41,共7页Proceedings of the CSU-EPSA

基　　金：国家自然科学基金资助项目(61102039);教育部新世纪优秀人才支持计划资助项目(NCET-11-0130);湖南省自然科学基金资助项目(14JJ7029)

摘　　要：为了提高变压器故障诊断的准确率和效率,合理评估变压器的状态,本文采用相关向量机RVM先对变压器的过热和放电故障进行划分。用自适应神经模糊推理系统进一步对故障进行分类,并对故障隶属概率进行估计。实验结果表明:本文方法有很强的学习能力和特征提取能力;尤其对于存在重叠区的故障特征,用模糊集和隶属度的方法能够输出故障类型概率,对状态评估进行决策辅助;诊断率高达96.15%,且运算效率高;跟支持向量机、人工神经网络方法相比,有更好的效率和更高的准确率。In order to improve the accuracy and efficiency of fault diagnosis for transformers and properly evaluate the corresponding statuses,relevance vector machine(RVM)is used to classify the overheating and discharge faults of transformers at first,then adaptive neural-fuzzy inference system(ANFIS)is utilized to identify the faults further,and estimate the probability of fault type.Experimental results show that the proposed method has a strong ability of learning and extracting features;especially,the method with fuzzy set and membership degree can output the probability of fault type in the case of overlapping fault features,which provides decision support for the status evaluation;the accuracy of the RVM-ANFIS method can reach as high as 96.15%,together with a higher calculation efficiency;compared with methods such as support vector machine(SVM)and artificial neural network(ANN),the proposed method has a better efficiency and a higher accuracy.

关 键 词：变压器状态评估 溶解气体分析 相关向量机 自适应神经模糊推理系统 故障诊断 

分 类 号：TM315[电气工程—电机]