基于粗糙集和模糊推理的高铁扼流适配变压器故障诊断研究  

Fault diagnosis of choke adapter transformer based on rough set and fuzzy inference for high-speed railway

作　　者：张楠乔 杨世武[2] 徐宗奇[1] 崔勇[2] ZHANG Nanqiao;YANG Shiwu;XU Zongqi;CUI Yong(Beijing National Railway Research & Design Institute of Signal & Communication Ltd., Beijing 100070, China;School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China)

机构地区：[1]北京全路通信信号研究设计院集团有限公司,北京100070 [2]北京交通大学电子信息工程学院,北京100044

出　　处：《铁路计算机应用》2017年第4期1-5,共5页Railway Computer Application

基　　金：中国铁路总公司技术标准项目(2016CR066);北京全路通信信号研究设计院集团有限公司委托项目资助(15KY-F-035)

摘　　要：文章梳理了扼流适配变压器故障模式和类型,完成了室内电路搭建及故障数据采集;选用粗糙集理论进行决策规则提取,在保持分类能力不变的条件下,去除监测数据表中大量的冗余特征,并提取出简明决策规则;进一步结合模糊推理,确定适用的隶属度,实现高正确率的故障诊断;最后,利用实际数据对此诊断方法的效果进行了验证,诊断正确率为92.5%,满足实际需求。This article combed failure modes and types of choke adapter transformer,implemented indoor circuit construction and fault data acquisition,chose the applicable fault diagnosis method,i.e.rough set theory,to obtain decision rules.On the condition that classification capability remained unchanged,plenty of redundancy features in the monitoring data table were reduced and concise decision rules were obtained.Forther more,this article used fuzzy inference to determine the applicable membership functions,implemented the fault diagnosis with high accuracy,used real data to verify the result of this fault diagnosis method.The accuracy was92.5%,it met the actual demands.

关 键 词：高速铁路 扼流适配变压器 故障诊断 粗糙集 模糊推理 

分 类 号：U224.2[交通运输工程—道路与铁道工程] TP39[自动化与计算机技术—计算机应用技术]