基于Petri网的离心式压缩机故障诊断研究

发布时间：2018-04-25 06:00

本文选题：故障诊断 + 离心式压缩机　；参考：《西南石油大学》2017年硕士论文

【摘要】：离心式压缩机是石油化工企业的关键设备,具有流向大,转速高,占地面积小等优点,承担着为整套设备和循环介质提供循环动力的任务。若离心压缩机产生故障,极其容易形成巨大的安全事故,给当事企业造成巨大危害。因此,如何在离心式压缩机出现故障征兆时尽可能快地采取措施规避故障事故的发生;如何在设备故障后快速找出原因并加以排除,将离心式压缩机故障事故降低到最低限度就成为了一个非常重要的研究课题。本文采用基于Petri网的故障诊断技术对离心式压缩机故障系统进行状态分析故障诊断。首先对基本Petri网的基础性知识和根本性规则进行定义和分析,接着再建立作用于故障信息传播演变推导分析的离心式压缩机故障Petri网系统模型,对离心式压缩机系统故障进行系统性推理分析,并利用模型的关联矩阵及其状态方程式分析法对故障信息的传播路径进行模拟论证。可是故障Petri网为具备模糊性以及不确定性的故障源进行反向诊断推导时具有严重缺陷,导致推导过程难以为继,对故障Petri网进行改进,使之能够符合故障诊断需要。为克服故障Petri网在推理分析复杂、不确定的故障信息中的不足,首先引入置信度最大及深度搜索优先的诊断办法,同时综合Petri网和模糊推导知识,形成并进一步定义出模糊Petri网诊断推理方法及其概念与规则表示,依照其反向诊断推理算法并基于已形成的故障来研判分析出故障源,同时分析罗列出该诊断推理方法的流程步骤,同时结合正向推理分析算法,通过逻辑推理和系统状态推理运算对离心式压缩机故障系统进行推导分析。经过对离心式压缩机系统进行正向推导分析及其反向诊断推导,得出了一致的结论,研判出系统中的故障源以及它们引发故障的可信度,由此论证了算法的有效性、可行性和准确性,进一步提升了故障诊断的准确性和高效性,为工作人员对离心式压缩机设备系统进行日常维护和故障诊断提供了一种新办法、新思路。
[Abstract]:Centrifugal compressor is the key equipment in petrochemical enterprises, which has the advantages of large flow direction, high speed, small area, and so on. It undertakes the task of providing circulating power for the whole equipment and circulation medium. If centrifugal compressor failure, it is easy to form a huge safety accident, causing great harm to the enterprise concerned. Therefore, how to take measures to avoid fault accidents as soon as possible, how to find out the cause quickly after the equipment failure and how to eliminate it. Reducing the fault of centrifugal compressor to the minimum has become a very important research topic. In this paper, fault diagnosis technology based on Petri net is used for fault diagnosis of centrifugal compressor fault system. Firstly, the basic knowledge and fundamental rules of basic Petri net are defined and analyzed, then the fault Petri net system model of centrifugal compressor is established, which is used to deduce and analyze the evolution of fault information. The fault of centrifugal compressor system is analyzed by systematic reasoning, and the propagation path of fault information is simulated by using the correlation matrix of the model and the analysis of the state equation. However, fault Petri nets have serious defects in the reverse diagnosis derivation for fault sources with fuzziness and uncertainty, which makes the derivation process difficult to continue. The fault Petri nets are improved to meet the needs of fault diagnosis. In order to overcome the shortcomings of fault Petri nets in reasoning and analysis of complex and uncertain fault information, a diagnosis method of maximum confidence and depth search priority is introduced, and Petri nets and fuzzy derivation knowledge are synthesized at the same time. The fuzzy Petri net diagnosis reasoning method and its concept and rule representation are formed and further defined. According to its reverse diagnosis reasoning algorithm and based on the fault that has been formed, the fault source is analyzed and analyzed. At the same time, the process steps of the diagnostic reasoning method are listed, and the fault system of centrifugal compressor is deduced and analyzed by logical reasoning and system state reasoning combined with forward reasoning analysis algorithm. After the forward derivation and reverse diagnosis of the centrifugal compressor system, a consistent conclusion is drawn, and the fault source and the reliability of the fault caused by the fault are studied, and the validity of the algorithm is demonstrated. The feasibility and accuracy improve the accuracy and efficiency of fault diagnosis, and provide a new method and new idea for the routine maintenance and fault diagnosis of centrifugal compressor equipment system.
【学位授予单位】：西南石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ051.21;TP301.1

【参考文献】