基于知识模型的城轨信号故障诊断系统  

An Urban Rail Signal Fault Diagnosis System Based on Knowledge Model

作　　者：杨将 阳亦斌 欧盛芬 邓永祁 YANG Jiang;YANG Yibin;OU Shengfen;DENG Yongqi(Hunan CRRC Times Signal&Communication Co.,Ltd.,Changsha,Hunan 410005,China)

机构地区：[1]湖南中车时代通信信号有限公司,湖南长沙410005

出　　处：《控制与信息技术》2022年第2期112-119,共8页CONTROL AND INFORMATION TECHNOLOGY

基　　金：系列化中国标准地铁列车研制及试验;长沙市科技重大专项(kh2003030)。

摘　　要：目前,城市轨道交通(简称“城轨”)信号维护支持系统仅能进行单一故障源报警,无法快速确定位故障原因和指导运维人员处置故障;而城轨信号系统故障种类繁多,诊断分析逻辑复杂,针对不同场景定制化开发故障诊断程序,无法快速响应运维需求且成本高昂。针对该问题,文章基于知识模型研发了一种信息化、平台化的城轨信号故障诊断系统,以实现信号系统故障知识建模、故障诊断语义关联和故障诊断过程建模。为了能够实现信号系统故障诊断的完全推理,系统采用网络本体语言的子语言(OWL DL)进行知识建模,抽取和描述信号系统故障诊断的分析逻辑并建立知识模型;利用设备运行状态匹配故障原因,由信号系统设备状态空间到故障原因空间的映射实现信号系统设备故障自诊断,为信号系统设备的生产运维管理提供决策支持。系统应用结果显示,采用该系统后运营安全事故率降低了15%,故障处置效率提升了20%。At present, urban rail signal maintenance system can only alarm a single fault source, and can not quickly locate the cause of fault and guide operation and maintenance personnel to deal with the fault. However, urban rail signal system has a large variety of faults, complex diagnosis and analysis logic, customized development of fault diagnosis procedures for different scenarios can not, quickly respond to the needs of operation and maintenance, and the cost is high. To solve this problem, this paper develops an information-based and platform-based urban rail signal fault diagnosis system based on knowledge model to realize signal system fault knowledge modeling, fault diagnosis semantic correlation and fault diagnosis process modeling. In order to realize the complete reasoning of signal system fault diagnosis, OWL DL(ontology web language description logic) is used to model knowledge, extract and describe the analysis logic of signal system fault diagnosis, and establish the knowledge model. The operation state of system equipment is used to match the fault cause, and the mapping from signal system equipment state space to fault cause space is used to realize the fault self-diagnosis of signal system equipment, so as to provide decision support for the production, operation and maintenance management of signal system equipment. Application results show that it can reduce the operation safety accident rate by 15% and improve the fault handling efficiency by 20%.

关 键 词：信号系统 知识模型 网络本体语言 故障诊断 Lambda架构 Hadoop技术生态圈 

分 类 号：U285[交通运输工程—交通信息工程及控制]