【作者】 张芳；

【导师】 黄敏；

【作者基本信息】 长沙理工大学 ， 通信与信息系统， 2012， 硕士

【摘要】 本文详细分析模糊Petri网与故障Petri网在故障诊断中的建模方法及对故障诊断的特点，发现这两类Petri网均存在不足，都不能完满的解决故障的诊断问题。其中模糊Petri网的运行规则不能体现故障传播的固有特性，且不能实现对故障的定位；而故障Petri网也存在不能确定故障发生的最大传播路径及反向推理时可能出现的冲突现象等问题。为此本文提出一种新的更适合于故障诊断的Petri网：模糊故障Petri网，并给出其定义、变迁使能及发生后果、故障推理算法。在Petri网的理论研究领域和故障诊断的应用领域都取得了较大的进步。在上述理论研究成果的基础上，采用基于UML的面向对象设计思想对该Petri网建模及推理算法进行实现。首先对Petri网推理系统原型进行需求分析，应实现功能为：文档/视图、模块编辑、建立模糊故障Petri网模型、正向推理及结果分析，进而生成系统原型的用例图，对各功能进行分配，获得其类图，完成分析模型向设计模型的转换。完成上述工作后，再对其进行具体编程实现，从而建立基于模糊故障Petri网建模方法的具有通用功能的推理系统原型。最后，本文将模糊故障Petri网故障诊断理论应用到风电机组的故障诊断中，再通过推理系统原型加以仿真，验证该方法的可行性和正确性，同时，为解决复杂环境下风电机组故障诊断和故障预警提供理论依据和实际方案。综合起来，本文的研究工作主要有以下几点：1、分析模糊Petri网与故障Petri网的建模方法在故障诊断中的特点，提出更适合故障诊断的模糊故障Petri网的概念。2、将模糊故障Petri网的动态特性与故障传播方式相结合，给出相应的模糊故障Petri网模型及诊断推理算法。3、使用Visual Studio工具研制模糊故障Petri网建模及推理系统的原型，建立库所、变迁及输入输出弧的对象集控件数据库，在界面上能实现对象添加、删除、拖动及修改，并能进行基本的推理。4、结合风电机组齿轮箱的故障传播机理，利用模糊故障Petri网建模方法对其进行建模和故障诊断，并运用所构建的系统原型对其进行仿真。更多还原

【Abstract】 This paper makes a detailed analysis of both the modeling methods and featuresof the Fuzzy Petri Net and Fault Petri Net in fault diagnosis. Deficiencies are foundthat neither of them can completely solve the fault diagnosis problems.For example,Fuzzy Petri Net’s operation rules don’t reflect the inherent characteristics of thefault propagation, neither can they locate the fault. However, Fault Petri Net can notascertain the maximum transmission path of the failure and there can emergeconflicts through backward reasoning. Based on this, Fuzzy Fault Petri Net is putforward, which is more suitable for fault diagnosis. Its definition, the transitionoccur rules and the consequences of the occurence,as well as the fault reasoningalgorithm are proposed.It makes a great deal of progress in the fields of Petri Nettheoretical research and fault diagnosis applications.Based on the above theoretical research results, the object-oriented ideas basedon UML is adopted to achieve the Petri net modeling and the inference algorithm. Inthe beginning, the requirement analysis of the Petri net reasoning system prototypeis made. The functions include editing documents, viewing modules, establishing themodel of the Fuzzy Fault Petri Net, and the forward reasoning and result analysis.Therefore, the use case diagram of the system prototype is generated. After thefunctions distrubed, the class diagram is gained to shift from analyzing model todesigning model. And then the reasoning system prototype with general functions isestablished.Finally, the theory of Fuzzy Fault Petri Net is applied to the fault diagnosis ofwind turbine.Then by using reasoning system prototype, it is verified that thismethod is feasibily and correct. At the same time, it provides a theoretical basis andpractical solutions to deal with the wind turbine fault diagnosis and fault warningunder the complex environment.Generally speaking, the study results of this paper are as follows:1. After analysis of Fuzzy Petri Net and Fault Petri Net modeling method in faultdiagnosis, the concept of Fuzzy Fault Petri Net which is more suitable for fault diagnosis is proposed.2.Combining the dynamic characteristics of Fuzzy Fault Petri Net and faultmode of transmission, the modeling method and diagnostic reasoning algorithm ofthe Fuzzy Fault Petri Net are given.3.By using Visual Studio tools to develop the prototype of the Fuzzy Fault Petrinet modeling method and the reasoning system, the place, transition and input/outputarc object sets of database are established, then the object is achieved to add, delete,modify and drag at the interface, and do the basic reasoning.4. With fault propagation mechanism of the wind turbine gearbox, the FuzzyFault Petri Net modeling approach is applied to model and fault diagnosis,what ismore, the constructed system prototype is used to simulate it.更多还原