工业过程中具有缓慢故障系统的可靠性预测研究

发布时间：2018-02-01 20:19

本文关键词： 系统可靠性预测缓慢故障贝叶斯理论数据驱动算法田纳西-伊斯曼过程　出处：《哈尔滨工业大学》2014年硕士论文　论文类型：学位论文

【摘要】：先进控制技术在过程控制中的应用不断加大，使得过程控制系统变得越来越复杂，由此带来系统故障种类的增多。对于某些工业过程来说，故障带来的损坏是永久的，无法修复或者修复系统的代价极其昂贵。若能掌控系统的工作状态并预测出系统可能发生故障的时间节点，及时更换相关故障环节的设备避免故障发生，就可以避免由于系统故障带来的经济损失，，因此预测出系统的可靠程度对系统的维护至关重要。一般的工业过程控制系统发生的故障可以分为两类分别为：突发性故障和缓慢故障。突发性故障是指系统由于不可测外力作用而导致的即时故障，随机性强，发生一般无征兆，因而一般不考虑其可靠性。缓慢故障是指系统由于自身设备磨损或元器件老化导致的延时性故障，系统正常工作时故障表现不明显，故而可以通过对系统易发生故障的部分进行实时监控来预测系统可能发生故障的节点。本文即针对这种故障来展开讨论的。本文通过四章来对系统的可靠性预测问题加以叙述。在第一章中对系统可靠性预测的研究现状作了简要的介绍，同时给出了本文所要研究的主要内容。第2章中给出了通过系统监控数据得到系统可靠性指标的方法，文中给出了两种方法，可以根据不同的数据使用不同的方法来得到系统可靠性指标。第3章提出了基于AR模型的贝叶斯预测方法，给出了具体的预测方案。第4章中对前面提出的方法基于TE过程的故障数据进行仿真，并最终得到系统可靠性的预测曲线。综上，本文给出了一种针对具有缓慢故障的工业过程控制系统进行可靠性预测的方法，将故障诊断于可靠性预测联系到一起，使得一些能用于故障诊断的方法应用到对系统可靠性预测领域当中。
[Abstract]:The application of advanced control technology in process control is increasing, which makes the process control system become more and more complex, resulting in an increase in the types of system failures. For some industrial processes. The damage caused by the failure is permanent, and it is extremely expensive to repair or repair the system. If you can control the working state of the system and predict the time node when the system may fail. It is very important to forecast the reliability of the system to avoid the economic loss caused by the system failure by replacing the equipment of the relevant fault link in time to avoid the failure. The common industrial process control system fault can be divided into two types: sudden failure and slow fault. Sudden fault refers to the system due to the role of undetectable external forces caused by the instant failure, strong randomness. Slow fault refers to the delay fault of the system caused by wear and tear of its own equipment or the aging of the components. The fault performance of the system is not obvious when it works normally. Therefore, it is possible to predict the possible nodes of the system by monitoring the vulnerable parts of the system in real time. In this paper, the reliability prediction of the system is described in four chapters. In the first chapter, the research status of the reliability prediction of the system is briefly introduced. At the same time, the main contents of this paper are given. In chapter 2, the method of obtaining system reliability index by system monitoring data is given, and two methods are given in this paper. According to different data, different methods can be used to obtain system reliability index. In chapter 3, Bayesian prediction method based on AR model is proposed. In chapter 4, the proposed method is simulated based on the fault data of te process, and the prediction curve of system reliability is obtained. In summary, this paper presents a method of reliability prediction for industrial process control system with slow fault, which connects fault diagnosis with reliability prediction. Some methods used in fault diagnosis can be applied to the field of system reliability prediction.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP273;TB114.3

【参考文献】