离散事件系统的可诊断性问题研究

发布时间：2018-10-23 18:28

【摘要】：2015年,“阿尔法”围棋战胜欧洲围棋冠军,人工智能成为了大家讨论的热门话题。随着人类社会的进步,人工智能逐渐成为我们生活中一个重要的领域。越来越多的智能设备为社会的发展而被应用到实际生产生活中,然而,一旦一个设备发生故障,就可能引起不可估量的损失和灾难。故障诊断是人工智能一个重要的研究分支。基于模型诊断是被广泛运用的故障诊断方法之一,该方法主要通过将系统的结构与行为模型以及实际行为与预期行为进行对比,从而推理出系统的诊断结果。离散事件系统的故障诊断近几年得到了越来越多学者的关注,本文围绕离散事件系统的故障诊断问题展开研究,主要贡献包括:首先,利用粗糙集理论解决离散事件系统中的故障诊断问题,用信息系统和决策系统表示离散事件系统,通过实际观测完善不完备模型,在决策系统中,提出了离散事件系统可诊断的充分必要条件,并基于粗糙集理论的性质,在不改变系统可诊断性的前提下,提出了极小化系统可观测事件的算法;其次,应用概率逻辑的方法诊断随机离散事件系统,提出了逻辑诊断器,通过逻辑诊断器和归结原理,判定随机离散事件系统的A-可诊断性和AA-可诊断性;最后,提出了分布式随机离散事件系统的故障诊断方法,并提出了同步随机诊断器,用于描述分布式随机离散事件系统A-可诊断的充分必要条件。本文主要创新点如下:利用粗糙集理论解决离散事件系统的故障诊断问题,有效提高了诊断效率;使用概率逻辑方法对随机离散事件系统中判定可诊断性的算法进行了改进;首次对分布式随机离散事件系统进行诊断问题的研究。
[Abstract]:In 2015, Alpha beat the European go champion and artificial intelligence became a hot topic. With the progress of human society, artificial intelligence has gradually become an important field in our life. More and more intelligent devices are applied to the actual production and life for the development of the society. However, once a device breaks down, it may cause incalculable losses and disasters. Fault diagnosis is an important branch of artificial intelligence. Model-based diagnosis is one of the widely used fault diagnosis methods. This method inferences the diagnosis results of the system by comparing the structure of the system with the behavior model and the actual behavior with the expected behavior. In recent years, more and more scholars have paid more and more attention to the fault diagnosis of discrete event systems. The main contributions of this paper are as follows: first of all, The problem of fault diagnosis in discrete event system is solved by rough set theory. The discrete event system is represented by information system and decision system. The sufficient and necessary conditions for the diagnosis of discrete event systems are proposed. Based on the properties of rough set theory and without changing the diagnosability of the systems, an algorithm for minimizing observable events is proposed. The method of probabilistic logic is used to diagnose the stochastic discrete event system. The logic diagnostics is put forward. The A- diagnosability and AA- diagnosability of the stochastic discrete event system are determined by logic diagnostics and resolution principle. In this paper, a fault diagnosis method for distributed stochastic discrete event systems is proposed, and a synchronous random diagnoser is proposed to describe the necessary and sufficient conditions for A- diagnosability of distributed stochastic discrete event systems. The main innovations of this paper are as follows: using rough set theory to solve the fault diagnosis problem of discrete event system effectively improves the diagnosis efficiency and improves the algorithm of judging diagnosability in stochastic discrete event system by using probabilistic logic method. The diagnosis of distributed stochastic discrete event systems is studied for the first time.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TP18;TP277

【相似文献】