基于模糊推理和粗糙集的旋转机械故障诊断的研究

发布时间：2018-06-29 20:29

本文选题：典型故障 + 框架理论　；参考：《江西理工大学》2011年硕士论文

【摘要】：故障诊断技术发展数十年来,已经从简单的观察,数据处理到以传感器和动态测试技术为手段,以信号处理和建模处理为基础的现代诊断技术,现在到了新的阶段——智能诊断技术阶段。本论文主要是以典型的旋转机械——风机为研究对象,在诊断系统构建方面做了相关研究。本文研究的重点是粗糙集约简可以让诊断过程更快速,也更准确;构建了诊断平台,使之能识别典型的故障,这之中构建了知识库和设计了知识推理的策略。系统来说,本文做了以下方面的研究和探讨: 1)介绍了故障征兆提取和故障推理方法。 2)用Access2003和框架理论建立风机征兆数据库,然后用VB6.0有效的连接建立起数据库,结合正向推理策略构建成诊断系统。 3)提取出各常见风机故障的特征,然后按照征兆顺序依次排列,一一对应数字化后,组成特征向量,运用粗糙集和约简理论,借助MATLAB实现证明约简后的诊断结果,更快,更全面。 4)采用了模糊推理理论,在建立知识库的基础上进行判别诊断。模糊推理是利用模糊性知识进行的一种不确定性推理,模糊推理理论的基础是模糊集理论以及在此基础上发展起来的模糊逻辑,它所处理的事物自身是模糊的,适用于故障知识的识别,通过建立起的仿真诊断也可说明这点。 5)运用VB6.0构建故障诊断系统时,当逻辑层次比较复杂的时候,无可避免的要面对复杂的知识搜索等问题,需要考虑知识组织和推理机制的改进,此系统具有不断完善的功能,知识库可以不断添加和维护,为此故障诊断可以随着使用时间的推移,诊断结果可以越来越精确。
[Abstract]:Over the past decades, fault diagnosis technology has developed from simple observation, data processing to modern diagnostic technology based on signal processing and modeling processing by means of sensors and dynamic test techniques. Now it's a new phase-intelligent diagnostic technology. In this paper, the typical rotating machinery-fan as the research object, in the diagnosis system construction has done the related research. The emphasis of this paper is that rough set reduction can make the diagnosis process faster and more accurate, and the diagnosis platform is constructed to identify typical faults, and the knowledge base and the strategy of knowledge reasoning are designed. In this paper, the following aspects are studied and discussed: 1) the methods of fault symptom extraction and fault reasoning are introduced. 2) the fan symptom database is established by using Access2003 and frame theory. Then the database is established with VB6.0 effectively, and the diagnosis system is constructed by combining forward reasoning strategy. 3) the features of the common fan faults are extracted, then arranged in sequence according to the symptoms, and then digitized one by one. The feature vector is made up of rough set and reduction theory, and the diagnosis result after the reduction is proved by MATLAB. 4) Fuzzy inference theory is adopted to discriminate diagnosis based on the establishment of knowledge base. Fuzzy reasoning is a kind of uncertain reasoning based on fuzzy knowledge. The basis of fuzzy reasoning theory is fuzzy set theory and fuzzy logic developed on this basis. It can also be illustrated by the simulation diagnosis built up. 5) when using VB6.0 to build fault diagnosis system, when the logic level is more complicated, It is necessary to consider the improvement of knowledge organization and reasoning mechanism in the face of complex knowledge search and other problems. This system has the function of continuous improvement, and the knowledge base can be added and maintained continuously. For this reason, fault diagnosis can be more and more accurate over time.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH165.3

【参考文献】