高炉故障诊断本体构建方法实证分析  

Empirical analysis on blast furnace fault diagnosis method based on ontology

作　　者：陈国荣[1,2] 鄢萍[2] 陈燕华[3] 施金良[1] 邓菊丽[1] 

机构地区：[1]重庆科技学院电气与信息工程学院,重庆401331 [2]重庆大学机械传动国家重点实验室,重庆400044 [3]重庆电力高等专科学校计算机系,重庆400053

出　　处：《重庆大学学报（自然科学版）》2012年第5期35-39,共5页Journal of Chongqing University

基　　金：国家'863'重点资助项目(2007AA040701);重庆市科技平台与基地建设资助项目(CSTC;2011PT-GC70006);重庆市教委成果转化重点项目(KJZH11221);重庆科技学院科研基金重点项目(CK2011B01;CK2010Z10;2011003)

摘　　要：高炉维修记录中包含很多有用的信息,可为诊断解决类似故障提供参考。针对故障记录提取和利用过程中存在的结构化表示、知识构建及推理难题,提出了一种基于本体的高炉故障知识构建及表示方法。对高炉故障诊断知识本体的特性进行分析,得出故障诊断本体构建的故障部位、故障原因、故障现象、解决方案四要素;定义了高炉本体的故障诊断类、属性、实例、关系、域等本体描述的规范框架和说明;在本体建模工具protégé中研究了该高炉本体的具体定义、描述和推理过程。实例研究表明,该方法简单、表述直观、效率高,可为高炉故障诊断提供一种新的思路和解决方案。Blast furnaces＇ maintenance records of mechanical fault contain great amount of information which is useful to identify the similar fault, but the structural representation and the knowledge reasoning problem prevent us to use the records effectively. Aiming at the problem,a blast furnace fault knowledge representation based on ontology is proposed. Blast furnaces＇ knowledge characteristics, ontology building significance and principle of fault are analyzed,and the four factors of BF-cause,BF-phenomenon,BF-ways, Blast-furnace in blast furnaces＇ mechanical fault diagnosis are brought out. The fault ontology frame of BFClass,BFProperties,BFIndividuals,BFRelationships,BFDomain are defined and described in detail. The applications of fault diagnosis ontology are studied in protege. The case study shows that the fault knowledge representation based on ontology is very intuitive and efficient,and it provides a novel way to deal with the blast furnace fault diagnosis problem.

关 键 词：诊断 本体 高炉 知识推理 

分 类 号：TH17[机械工程—机械制造及自动化]