基于知识图谱的信号设备故障诊断方法  

Fault Diagnosis Method of Signal Equipment Based on Knowledge Graph

作　　者：王晴雯 张振海[1] WANG Qingwen;ZHANG Zhenhai(School of Automation and Electrical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)

机构地区：[1]兰州交通大学自动化与电气工程学院,兰州730070

出　　处：《铁道标准设计》2022年第12期149-156,共8页Railway Standard Design

基　　金：国家自然科学基金项目(61763025);中国博士后科学基金资助项目(167306)。

摘　　要：信号设备故障处理的大部分知识都分布于专家经验、运营规程、运营作业指导书等零散的故障分析记录里,既不便于对大量的故障分析语料进行管理,也不易于操作人员进行查询与学习。针对列控系统信号设备故障分析的零散性与复杂性,提出一种结合自顶向下和自底向上方式来构建信号设备故障知识图谱的构建方法;通过知识抽取、知识存储、知识图谱展示,实现信号设备故障分析语料的可视化展现;同时,提出基于Cypher语言的信号设备故障维修方案查询方法。结果表明:构建的信号设备故障知识图谱能够有效实现故障数据的结构化存储与可视化展现,提出的故障维修方案查询方法能够实现故障概况的相应匹配。构建信号设备知识图谱提高了故障数据的数字化管理水平,有助于在信号设备发生故障时做出辅助决策推理。Most knowledge of signal equipment fault processing is distributed in the scattered fault analysis records such as expert experience,operation procedures,operation instructions,and etc.This is neither convenient to manage a large number of fault analysis data,nor easy for operators to inquire and learn.Thus,a method combining top-down and bottom-up modes to construct fault knowledge graph of signal equipment is proposed given the mentality and complexity of signal equipment fault analysis in train control system.Through knowledge extraction,knowledge storage and knowledge graph display,the visual display of signal equipment fault analysis corpus is realized.At the same time,a query method of signal equipment fault maintenance scheme based on Cypher language is proposed.The results show that the constructed fault knowledge graph can effectively realize the structured storage and visualization of fault data,and the proposed fault maintenance scheme query method can fulfill corresponding matching of fault profile.The construction of signal equipment knowledge graph improves the digital management level of fault data,and helps to make auxiliary decision reasoning in case of signal equipment failure.

关 键 词：知识图谱 信号设备故障 知识存储 图数据库 辅助决策 

分 类 号：U284.92[交通运输工程—交通信息工程及控制]