基于改进Rete算法的旋转机械故障诊断专家系统的研究

发布时间：2018-06-16 09:44

本文选题：旋转机械 + 故障诊断　；参考：《北京化工大学》2011年硕士论文

【摘要】：自工业革命以来,社会生产水平得到飞速的发展,为了满足人们对产品日益增长的需求,工业自动化生产水平不断提高。在这种背景下,生产设备日趋向大型化、复杂化、自动化、连续性生产方向发展。一旦工厂关键机组出现故障停机,往往造成重大安全事故和严重经济损失,特别是在石化企业,造成的损失往往数以百万计。因此,如何保证生产线能够安全稳定的运行就成了各大企业一个非常重要的课题,设备故障诊断技术也是在这一需求背景下发展而来的。在工业生产中,应用先进的设备故障诊断技术可以有效准确的对设备进行故障诊断,分析其故障机理与故障原因,从而有效改善设备维修效率,提高企业经济效益,降低设备维护成本并减少企业发生重大安全事故几率,因此具有重要的经济意义和应用价值。本论文以旋转机械智能故障诊断技术为研究对象,结合专家系统技术,对旋转机械智能故障诊断技术进行了研究,全文主要内容主要分为如下几部分。第一章绪论部分主要介绍本课题来源和研究意义,并简单概述了故障诊断技术的发展历史和专家系统在故障诊断领域的应用现状,以及当前智能故障诊断技术的发展方向和前景,并在最后提出本课题的研究内容和研究目的。第二章主要针对本课题的研究对象,系统阐述了专家系统领域的经典前向匹配算法Rete算法的发展历史以及算法数据流网络实现原理,并根据本课题需求用伪代码的形式描述了该算法的具体实现。第三章主要针对旋转机械常见故障包括转子动平衡、转子不对中、转子弯曲等常见故障的特征和机理进行深入分析,并在最后将旋转机械常见故障以表格的形式进行了整理,这也是本课题专家系统知识库的建立基础。旋转机械本身结构复杂,且在现场经常处理连续工作状态,因此其故障具有多样性和复杂性的特点,同一种故障常常会表现出多种征兆,一种征兆也往往是多种故障的叠加的结果。因此对旋转机械常见故障进行分析整理具有重要的实际意义, 第四章在前文内容基础上,利用前向快速匹配算法Rete算法作为专家系统规则推理算法,以第三章所总结的故障征兆表作为基础整理的知识规则库,利用Java语言和Eclipse开发平台设计并完成旋转机械故障诊断专家系统。第五章对本课题的完成情况进行了回顾,并指出了本课题研究过程中的缺陷以及对下一步研究工作的展望与建议。
[Abstract]:Since the Industrial Revolution, the level of social production has developed rapidly. In order to meet the increasing demand for products, the level of industrial automation production has been improved. In this context, production equipment tends to be large, complex, automatic, continuous production direction. Once the critical units of the plant are shut down, they often cause serious safety accidents and serious economic losses, especially in petrochemical enterprises, which often result in millions of losses. Therefore, how to ensure the safe and stable operation of the production line has become a very important issue for the large enterprises, and the equipment fault diagnosis technology has been developed under the background of this demand. In industrial production, the application of advanced equipment fault diagnosis technology can effectively and accurately diagnose the equipment, analyze its fault mechanism and cause, thus effectively improve the equipment maintenance efficiency and increase the economic benefit of the enterprise. It has important economic significance and application value to reduce the cost of equipment maintenance and reduce the probability of major safety accidents in enterprises. This paper takes the intelligent fault diagnosis technology of rotating machinery as the research object and combines the expert system technology to study the intelligent fault diagnosis technology of rotating machinery. The main content of this paper is divided into the following parts. The first chapter introduces the origin and research significance of this topic, and briefly summarizes the history of fault diagnosis technology and the application of expert system in fault diagnosis field. And the development direction and prospect of intelligent fault diagnosis technology at present, and finally put forward the research content and research purpose of this subject. In the second chapter, aiming at the research object of this subject, the development history of the classical forward matching algorithm Rete algorithm and the realization principle of the algorithm data stream network in the field of expert system are systematically expounded. The implementation of the algorithm is described in the form of pseudo code according to the requirements of this paper. The third chapter mainly analyzes the characteristics and mechanism of the common faults of rotating machinery, such as rotor dynamic balance, rotor misalignment, rotor bending and so on. At last, the common faults of rotating machinery are sorted out in the form of table. This is also the foundation of the expert system knowledge base. The rotating machinery itself is complex in structure and often deals with the continuous working state in the field, so its faults have the characteristics of diversity and complexity, the same kind of faults often show many kinds of symptoms. A symptom is also often the result of a combination of failures. Therefore, it is of great practical significance to analyze and sort out the common faults of rotating machinery. In chapter 4, the forward fast matching algorithm, Rete algorithm, is used as the expert system rule reasoning algorithm based on the previous contents. Based on the fault symptom table summarized in Chapter 3, a fault diagnosis expert system for rotating machinery is designed and completed by using Java language and Eclipse development platform. The fifth chapter reviews the completion of the project and points out the defects in the research process and the prospects and suggestions for the next research work.
【学位授予单位】：北京化工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3;TP182

【参考文献】