基于多GEP分类器和DGA的变压器故障诊断  

Transformer fault diagnosis based on multi-GEP classifier and DGA

作　　者：牛中敏 牛继涛[2] 董卓[3] 

机构地区：[1]河南省电力勘测设计院,河南郑州450007 [2]河南郑州供电公司,河南郑州450006 [3]华北电力大学电子与电气工程学院,河北保定071003

出　　处：《华北电力大学学报（自然科学版）》2012年第3期35-40,共6页Journal of North China Electric Power University：Natural Science Edition

摘　　要：将GEP(Gene Expression Programming)方法与变压器油中溶解气体分析方法结合起来,提出了基于自适应GEP分类算法的变压器故障诊断方法。该方法继承了遗传算法(GA)的线性性和遗传程序设计(GP)的普适性,从而达到了简单编码解决复杂问题的目的,具有良好的收敛性和鲁棒性。选择能反映各种故障而又不冗余的400组DGA实测数据作为GEP分类器的训练样本和测试样本,并将测试结果与NB分类器,BP网络法,免疫分类法进行对比分析。大量诊断实例表明,所提出的自适应多GEP分类方法适用于变压器故障诊断,其性能优于另外3种方法。GEP （Gene Expression Programming） method is combined with transformer oil dissolved gas analysis, and also a method of transformer fault diagnosis based on self-adaptive GEP classification algorithm is proposed. Because of the combination of GEP method has inherited the linear character of genetic algorithm （GA） and the universal character of genetic programming （GP） , the method achieves the purpose of solving complex problems with simple codes, and also has good convergence and robustness. 400 groups of DGA measured data are chosen as the training samples and test samples of the GEP classifier, and the samples include a variety of failure and are non-redundant. Test results of the proposed method are compared with that of the NB classifier, BP network method, and the immune classification. A large number of diagnostic examples show that the proposed self-adaptive classification GEP is suitable for transformer fault diagnosis, and its performance is better than the other three methods.

关 键 词：变压器故障诊断 基因表达式程序设计 DGA 数据归一化 

分 类 号：TM4[电气工程—电器]