基于蜻蜓算法和支持向量机的变压器故障诊断  

Transformer Fault Diagnosis Based on Dragonfly Optimization Algorithm and Support Vector Machine

作　　者：傅军栋[1] 陈俐[1] 康水华 冯奕轩 

机构地区：[1]华东交通大学电气与自动化工程学院,江西南昌330013

出　　处：《华东交通大学学报》2016年第4期103-112,共10页Journal of East China Jiaotong University

摘　　要：电力变压器作为电力系统中变换电压、输送电能、分配电能的重要电力设备,其运行状态直接影响到电网的运行安全。为了提高故障诊断的准确率,提出了一种基于蜻蜓算法(dragonfly algorithm,DA)和支持向量机(support vector machine,SVM)的变压器故障诊断方法。利用蜻蜓算法DA对SVM分类器参数进行优化且以SVM分类预测准确率最大为其适应度函数。变压器故障诊断实例仿真结果分析表明:基于蜻蜓算法的DA-SVM支持向量机的变压器故障诊断与交叉验证法CV-SVM,基于遗传算法的GA-SVM,基于粒子群算法的PSO-SVM相比较,在变压器故障诊断中具有故障诊断准确率高,全局寻优能力强,收敛速度快,且稳定性好的优越性。Power transformer is an important electrical equipment used to transform voltage,transmit and distribute power,and its running state directly affects the safety of power network. In order to improve the accuracy of fault diagnosis,this paper puts forward a kind of method for transformer fault diagnosis based on Dragonfly Algorithm（DA） and support vector machine（SVM）. Parameters of SVM classifier are optimized by DA and fitness function is implemented according to the maximum accuracy of SVM classification prediction. Transformer fault diagnosis example results show that,DA-SVM based on dragonfly algorithm,compared with cross validation method of transformer fault diagnosis CV-SVM,GA-SVM based on genetic algorithm（GA）,and PSO-SVM based on particle swarm algorithm,has high accuracy of fault diagnosis,good global search capability,high convergence speed and superiority with good stability.

关 键 词：蜻蜓算法 支持向量机 变压器 故障诊断 

分 类 号：TM715[电气工程—电力系统及自动化]