地铁车辆大功率逆变器故障诊断研究  

Study of Fault Diagnosis of Metro Vehicle High-power Inverters

作　　者：李小波[1] 秦文[1] 王泉 

机构地区：[1]上海工程技术大学城市轨道交通学院,上海201620 [2]上海地铁电子科技有限公司,上海200233

出　　处：《计算机仿真》2016年第7期229-233,共5页Computer Simulation

基　　金：上海市研究生教育创新计划学位点引导布局与建设培育项目(13sc002)

摘　　要：地铁车辆大功率逆变器故障直接影响车辆安全可靠运行,甚至带来救援、退出运营的后果。目前的诊断方法只适用于单个故障识别,对多个故障识别时特征数据提取复杂。提出了基于支持向量机(SVM)的逆变器故障诊断方法。通过Clark变换将三相坐标系转换为两相坐标系,计算在两相坐标系中合成的空间矢量模值,利用不同故障类型情况下空间矢量模值的畸变位置不同来提取故障特征向量,再通过支持向量机对故障特征向量的分类来实现对不同故障类型的诊断。SVM方法能有效地诊断出单个功率管故障和两个功率管同时故障的故障类型,在MATLAB中的仿真验证了SVM方法的准确性,对地铁车辆大功率逆变器的维修维护具有重要的参考意义。The faults of metro vehicle high power inverter directly affect the vehicle operating safety and reliabity, even bring evacuation or withdraw from the operation of the consequences. At present, the fault diagnosis method is only suitable for single fault identification, and it is complex to extract feature data about several fault identification. A fault diagnosis method based on support vector machine （SVM） is proposed in this paper. Through converting the three-phase coordinate system into two-phase system by Clark transformation, and calculating synthetic space vector modulus which are various from different fault types in the two-phase system, the space veetor modulus is used to extract fault characteristics and the SVM is used to isolate fault modes. This method is effectively to diagnose such fault types as one switch fault or two switch faults at the same time, and the simulation in MATLAB verifies its accuracy.

关 键 词：地铁车辆 大功率逆变器 故障诊断 支持向量机 空间矢量 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]