基于本体的汽轮发电机组故障诊断知识建模  

Knowledge modeling of fault diagnosis for turbine generator sets based on ontology

作　　者：剡昌锋[1] 栗宇 王慧滨 张强 艾科勇 吴黎晓[1] YAN Chang-feng;LI Yu;WANG Hui-bin;ZHANG QiangAI Ke-yong;WU Li-xiao(College of Mechano-Electronic Engineering, Lanzhou Univ. of Tech., Lanzhou 730050, China;Zhangzhou Health Vocational College, Zhangzhou 363000, China;Guangzhou Locomotive Co. Ltd., Guangzhou 510830, China)

机构地区：[1]兰州理工大学机电工程学院,甘肃兰州730050 [2]漳州卫生职业学院,福建漳州363000 [3]广州电力机车有限公司,广东广州510830

出　　处：《兰州理工大学学报》2020年第5期41-48,共8页Journal of Lanzhou University of Technology

基　　金：国家自然科学基金(51765034,51165018)。

摘　　要：针对目前汽轮发电机组故障诊断领域知识术语复杂、系统异构、知识表示不完备以及共享和重复使用困难等问题,依据故障诊断需求,采用基于本体的知识表示方法,提出了一种适用于汽轮发电机组故障诊断领域的本体构建方法和知识表示模型.在解析了汽轮发电机组故障知识特性的前提下,定义了其本体概念、属性、关系、实例和公理,为知识表示提供了明确的形式化规格说明,并借助Protégé_4.3构建了包含汽轮发电机组的故障类型、故障特征、故障原因和维修策略等故障诊断领域本体,设计了一致性检验的算法.在此基础上,在SQI机械故障综合模拟实验台上模拟汽轮发电机组故障,通过FaCT++推理机实现本体知识推理测试.结果表明基于本体的汽轮发电机组故障诊断知识模型是可行的.Since the knowledge representation method of fault diagnosis in turbine generator sets is lack of complete in the field of terms,complex terminology,system heterogeneity and difficulty in sharing and reuse etc,a new method for ontology-based knowledge representation is well adopted in terms of fault diagnosis requirements.Methods for ontology construction and a knowledge representation model for the fault diagnosis of turbine generator sets are proposed respectively.The model defines its ontology concepts,attributes,relationships,examples and axioms,and provides a clear formal specification for knowledge representation.Furthermore,Protégé_4.3 is used to construct the ontology of turbine generator sets fault diagnosis domain with fault type,fault characteristic,fault reason and maintenance strategy,and the knowledge is proved to be consistent in the ontology by designing algorithms.Turbine generator sets fault is simulated by SQI mechanical fault comprehensive simulation test bench,and the reasoning test of ontology knowledge is verified by the inference engine named FaCT++.The result indicates that the knowledge model of ontology-based fault diagnosis is feasible.

关 键 词：汽轮发电机组 故障诊断 知识建模 一致性检验 本体推理 

分 类 号：TH17[机械工程—机械制造及自动化] TP182[自动化与计算机技术—控制理论与控制工程]