飞行器故障诊断与辅助决策技术

发布时间：2018-06-05 09:40

本文选题：动态故障树 + 静态故障树　；参考：《西安工业大学》2016年硕士论文

【摘要】：故障诊断技术是一项与生产现实具有密切关系的工程类学科,广泛应用于当前工业、国防等领域,是一个日益重要的研究内容。随着系统复杂程度的越来越高,以及自动化程度的快速发展,系统故障率也随之升高,对系统的可靠性要求也越来越高。所以,在未来很长时间内,故障诊断技术必定会成为国内外专家学者们的重要研究方向。利用先进的故障诊断技术对系统进行诊断,不仅可以有效排除系统故障,而且可以防止重大事故的发生,避免人员伤亡及经济损失。本文以航空航天领域为背景,深入研究了“飞行器故障诊断技术”的理论方法和实际应用实例。主要针对如下两项工作做了大量的研究和开发：一是进行了基于故障树和有向图的故障诊断算法研究,并分别应用该方法进行了飞行器实际系统的故障诊断；二是基于C#软件开发平台完成了“飞行器故障诊断与辅助决策系统”的部分功能开发。以下是对本课题内容和成果的归结。(1)静态故障树方法研究。包括静态故障树定性、定量分析；采用二元决策图进行静态故障树分析。(2)动态故障树方法研究。第一,提出了基于Monte Carlo的动态故障树顶事件发生概计算方法。对于包含不同动态逻辑门的动态故障树模型,将顶事件发生概率转换为不同逻辑门的多重积分,采用Monte Carlo近似方法计算,将此方法应用于“某卫星太阳翼驱动机构步进电机”故障诊断实例；第二,把模块化的思想引入动态故障树进行分析。采用此方法首先需要将动态故障树模型分解成动态和静态子树,再各自采取Markov模型和二元决策图法进行求解,完成了“某卫星太阳翼-蓄电池组光伏电源系统”的故障分析。(3)有向图的故障分析方法。第一,研究通过分层SDG模型进行故障诊断的方法。采用分层策略,缩小故障源搜索空间,根据测量节点之间的内在联系向前搜索,判断是否为相容支路,获得备选故障源的集合。第二,将模糊思想和分层SDG模型相结合,实现了模糊分层SDG模型故障分析方法。该方法首先需要对系统进行SDG建模,而后进行分层,采用模糊变量来表示节点变量,采取CPT来描述各节点之间的内在联系,采用贝叶斯向前寻找的方法把所有备选故障源找到,对此全部备选故障源进行可能性大小排列。利用上述两种方法分别针对“某型号航空发动机燃油调节系统”进行了故障诊断。(4)基于C#软件开发平台完成了“飞行器故障诊断与辅助决策系统”部分功能开发,实现了静态故障树绘制。
[Abstract]:Fault diagnosis technology is an engineering subject closely related to production reality. It is widely used in industry, national defense and other fields. It is an increasingly important research content. With the increasing complexity of the system and the rapid development of automation, the failure rate of the system also increases, and the reliability requirements of the system become higher and higher. Therefore, in a long time in the future, fault diagnosis technology will become an important research direction of experts and scholars at home and abroad. Using advanced fault diagnosis technology to diagnose the system can not only effectively eliminate the system failures but also prevent the occurrence of serious accidents and avoid casualties and economic losses. In this paper, based on the background of aerospace field, the theoretical method and practical application of "aircraft fault diagnosis technology" are studied in depth. A lot of research and development have been done for the following two tasks: first, the fault diagnosis algorithm based on fault tree and directed graph is studied, and the fault diagnosis method is applied to the actual flight vehicle system respectively; Secondly, based on C # software development platform, some functions of aircraft fault diagnosis and auxiliary decision system are developed. The following is the research of static fault tree method. It includes the qualitative and quantitative analysis of static fault tree and the research of dynamic fault tree method by using binary decision graph to analyze static fault tree. Firstly, a method for calculating the occurrence of dynamic fault tree top events based on Monte Carlo is proposed. For the dynamic fault tree model with different dynamic logic gates, the probability of top event occurrence is converted into multiple integrals of different logic gates, and the Monte Carlo approximation method is used to calculate it. The method is applied to the fault diagnosis example of the stepping motor of a satellite solar wing drive mechanism. Secondly, the modularization idea is introduced into the dynamic fault tree for analysis. Using this method, the dynamic fault tree model should be decomposed into dynamic and static subtrees, and then Markov model and binary decision graph method should be used to solve the problem. The fault analysis method of a satellite solar wing-battery photovoltaic power supply system with digraph is completed. First, the method of fault diagnosis based on hierarchical SDG model is studied. The hierarchical strategy is adopted to narrow down the search space of fault sources and to search forward according to the internal relations between measurement nodes to determine whether they are compatible branches and to obtain the set of alternative fault sources. Secondly, the fault analysis method of fuzzy hierarchical SDG model is realized by combining fuzzy idea with hierarchical SDG model. This method first needs to model the system with SDG, then stratifies, uses fuzzy variables to represent node variables, adopts CPT to describe the internal relations between nodes, and uses Bayes forward search method to find all alternative fault sources. Arrange the possible size of all the alternative fault sources. Using the above two methods, the fault diagnosis of a certain type of aero-engine fuel regulation system is carried out respectively. (4) based on the C # software development platform, the partial function of "aircraft Fault diagnosis and Auxiliary Decision-Making system" is developed. The static fault tree drawing is realized.
【学位授予单位】：西安工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V267

【相似文献】