机械故障智能诊断系统本体建模及推理的应用研究
发布时间：2019-03-03 20:40
【摘要】：近年来,机械故障智能诊断系统的研究成为机械故障诊断领域的研究热点,而系统的智能化程度和诊断准确度依赖于系统知识库中知识的数量与质量以及知识的组织、分类以及更进一步的知识共享和推理。本体作为一种能在语义和知识层次上描述知识模型的建模工具,可以提供对某一领域的概念以及相互关系的概念化描述,为知识共享奠定基础,本体支持对知识信息的区分,可以实现对领域知识的层次化表示,本体中包含的类公理以及约束公理可以用于知识的推理,且现有的本体推理机可以对本体提供直接的推理服务。 本文主要研究如何应用本体构建机械故障智能诊断系统的知识库,使用本体推理保证知识库知识的正确性以及通过本体推理实现故障诊断的方法。具体对以下内容进行了分析和研究: (1)针对机械故障诊断需求,在现有的领域本体建模方法的基础上提出了一套适用于机械故障诊断领域的本体构建方法,通过对故障领域知识的分析,提出了纵向建模方案,对故障诊断领域知识根据故障类型、故障征兆、故障原因以及故障处理方法进行分类。通过定义属性来表达故障类型、故障征兆、故障原因以及故障处理方法之间存在的复杂的对应关系,从而将故障诊断领域知识组织为具备语义关系的本体知识库。 (2)依据提出的机械故障诊断领域的本体构建方法以及纵向建模方案,以一类交流电动机故障为例,使用本体建模工具Protégé4.0构建了交流电动机故障诊断本体,验证了本文提出的方法的可行性。 (3)本文提出通过定义复合类的方法实现故障诊断,即使用OWL中的构造子对故障类型做复合类定义并将其声明为等价类,本体推理机通过对等价类的推理实现故障诊断。 (4)使用本体推理机对故障诊断本体进行了推理,对于推理检测出的逻辑错误提出了相应的解决方案,并通过对等价类的推理,实现了对起动故障、过热故障以及电刷等故障的诊断,其诊断结果的准确性和合理性较高。 结果表明,本文提出的机械故障诊断领域的本体构建及推理的方法具有可行性,为机械故障智能诊断领域的本体建模及推理的理论与方法研究提供了一条值得探索的途径。
[Abstract]:In recent years, the research of mechanical fault intelligent diagnosis system has become a hot topic in the field of mechanical fault diagnosis, and the degree of intelligence and diagnostic accuracy of the system depends on the quantity and quality of knowledge and the organization of knowledge in the knowledge base of the system. Classification and further knowledge sharing and reasoning. Ontology, as a modeling tool that can describe the knowledge model at the semantic and knowledge level, can provide a conceptual description of the concepts and relationships in a certain domain, lay the foundation for knowledge sharing, and ontology supports the distinction of knowledge information. The hierarchical representation of domain knowledge can be realized. The class axiom and constraint axiom contained in ontology can be used for reasoning of knowledge, and the existing ontology inference machine can provide direct reasoning services to ontology. This paper mainly studies how to use ontology to construct the knowledge base of mechanical fault intelligent diagnosis system, and how to use ontology reasoning to ensure the correctness of knowledge base and the method of realizing fault diagnosis through ontology reasoning. The following contents are analyzed and studied in detail: (1) according to the requirements of mechanical fault diagnosis, a set of ontology construction methods suitable for mechanical fault diagnosis is proposed on the basis of existing domain ontology modeling methods. Based on the analysis of fault domain knowledge, a vertical modeling scheme is proposed. The fault diagnosis domain knowledge is classified according to the fault type, fault symptom, fault cause and fault handling method. By defining attributes to express the complicated correspondence among fault type, fault symptom, fault cause and fault handling method, the knowledge of fault diagnosis field is organized into ontology knowledge base with semantic relationship. (2) according to the ontology construction method and longitudinal modeling scheme in the field of mechanical fault diagnosis, the fault diagnosis ontology of AC motor is constructed by using the ontology modeling tool Prot 茅 g 茅 4.0, taking a class of AC motor faults as an example. The feasibility of the proposed method is verified. (3) this paper proposes a method to realize fault diagnosis by defining composite class. Even if the composite class is defined by the constructors in OWL and declared as equivalent class, the ontology inference machine realizes fault diagnosis by reasoning the equivalent class. (4) the ontology inference machine is used to reason the fault diagnosis ontology, and the corresponding solution to the logic error detected by reasoning is put forward, and the starting fault is realized by reasoning the equivalent class. The diagnosis of overheating fault and brush fault has high accuracy and rationality. The results show that the method of ontology construction and reasoning in the field of mechanical fault diagnosis proposed in this paper is feasible, which provides a valuable way to study the theory and method of ontology modeling and reasoning in the field of intelligent mechanical fault diagnosis.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TP391.1;TH165.3
本文编号：2434078
[Abstract]:In recent years, the research of mechanical fault intelligent diagnosis system has become a hot topic in the field of mechanical fault diagnosis, and the degree of intelligence and diagnostic accuracy of the system depends on the quantity and quality of knowledge and the organization of knowledge in the knowledge base of the system. Classification and further knowledge sharing and reasoning. Ontology, as a modeling tool that can describe the knowledge model at the semantic and knowledge level, can provide a conceptual description of the concepts and relationships in a certain domain, lay the foundation for knowledge sharing, and ontology supports the distinction of knowledge information. The hierarchical representation of domain knowledge can be realized. The class axiom and constraint axiom contained in ontology can be used for reasoning of knowledge, and the existing ontology inference machine can provide direct reasoning services to ontology. This paper mainly studies how to use ontology to construct the knowledge base of mechanical fault intelligent diagnosis system, and how to use ontology reasoning to ensure the correctness of knowledge base and the method of realizing fault diagnosis through ontology reasoning. The following contents are analyzed and studied in detail: (1) according to the requirements of mechanical fault diagnosis, a set of ontology construction methods suitable for mechanical fault diagnosis is proposed on the basis of existing domain ontology modeling methods. Based on the analysis of fault domain knowledge, a vertical modeling scheme is proposed. The fault diagnosis domain knowledge is classified according to the fault type, fault symptom, fault cause and fault handling method. By defining attributes to express the complicated correspondence among fault type, fault symptom, fault cause and fault handling method, the knowledge of fault diagnosis field is organized into ontology knowledge base with semantic relationship. (2) according to the ontology construction method and longitudinal modeling scheme in the field of mechanical fault diagnosis, the fault diagnosis ontology of AC motor is constructed by using the ontology modeling tool Prot 茅 g 茅 4.0, taking a class of AC motor faults as an example. The feasibility of the proposed method is verified. (3) this paper proposes a method to realize fault diagnosis by defining composite class. Even if the composite class is defined by the constructors in OWL and declared as equivalent class, the ontology inference machine realizes fault diagnosis by reasoning the equivalent class. (4) the ontology inference machine is used to reason the fault diagnosis ontology, and the corresponding solution to the logic error detected by reasoning is put forward, and the starting fault is realized by reasoning the equivalent class. The diagnosis of overheating fault and brush fault has high accuracy and rationality. The results show that the method of ontology construction and reasoning in the field of mechanical fault diagnosis proposed in this paper is feasible, which provides a valuable way to study the theory and method of ontology modeling and reasoning in the field of intelligent mechanical fault diagnosis.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TP391.1;TH165.3
【参考文献】
相关期刊论文 前10条
1 邓志鸿,唐世渭,张铭,杨冬青,陈捷;Ontology研究综述[J];北京大学学报(自然科学版);2002年05期
2 肖洪;余锦凤;;数字图书馆领域本体构建研究——以数字参考咨询领域为例[J];大学图书馆学报;2006年06期
3 石莲;孙吉贵;;描述逻辑综述[J];计算机科学;2006年01期
4 何庆;汤庸;黄永钊;;基于本体的法律知识库的研究与实现[J];计算机科学;2007年02期
5 王进鹏;张亚非;苗壮;;基于本体的关系数据集成的查询处理[J];计算机科学;2010年12期
6 雍兮;曹存根;白埃民;;基于本体的智能规划方法研究及应用[J];计算机科学;2011年02期
7 杜磊;易建军;季白杨;董金祥;余斌;;应用本体构造发动机故障诊断专家系统[J];计算机应用;2006年S2期
8 窦丹丹;谢强;丁秋林;;基于Ontology的故障诊断方法[J];计算机应用;2009年02期
9 王辉;冯志勇;陈炬;陈世展;;基于本体的服务网络平台及其构建方法[J];计算机应用;2010年08期
10 李广明;;基于本体的知识资源模糊聚类分析[J];计算机应用研究;2011年02期
相关硕士学位论文 前2条
1 汪智勇;本体查询与推理研究及其实现[D];中南大学;2007年
2 孙鹏;语义网中基于描述逻辑的本体推理研究[D];吉林大学;2009年
，本文编号：2434078
本文链接：https://www.wllwen.com/kejilunwen/jixiegongcheng/2434078.html
