特种车辆故障知识图谱构建方法研究  

Research on the Construction Method of Fault Knowledge Graph for Special Vehicles

作　　者：苗凤金 赵金龙[1] 柳月[1] 王秋芳[1] 杨立煜 MIAO Fengjin;ZHAO Jinlong;LIU Yue;WANG Qiufang;YANG Liyu(China North Vehicle Research Institute,Beijing 100072,China)

机构地区：[1]中国北方车辆研究所,北京100072

出　　处：《火力与指挥控制》2025年第2期177-181,共5页Fire Control & Command Control

摘　　要：针对由于特种车辆故障信息利用率低导致维修效率低的问题,结合知识图谱相关技术,提出一种用于特种车辆的故障知识图谱构建方法。将特种车辆已有的故障信息作为输入,采用BERT-BiLSTM-CRF模型对故障信息进行知识提取,对提取后结构化的故障信息进行信息融合,以三元组的形式存储于Neo4j图数据库中,最终服务于故障智能问答和故障推理等故障相关知识应用。特种车辆故障知识图谱的构建和应用将大幅提高其信息利用率及维修性,为其在战场上抢占先机提供了助力。Aiming at the problem of low maintenance efficiency caused by low utilization rate of special vehicle fault information,a method of fault knowledge graph construction for special vehicles is proposed based on the related technology of knowledge graph.First,the existing fault information of special vehicles is taken as input,and the BERT-BiLSTM-CRF model is used to extract the fault information.Then,the extracted structured fault information is fused and stored in the Neo4j graph database in the form of triplet.Finally,the finished knowledge graph can be utilized to serve such applications as fault intelligent question-answering system and fault reasoning system,etc.The construction and application of special vehicle fault knowledge graph will greatly improve its information utilization rate and maintainability,which provides a boost for the special vehicle to seize the opportunity in the battlefield.

关 键 词：特种车辆 故障信息 知识图谱 维修性 

分 类 号：TJ811[兵器科学与技术—武器系统与运用工程]