大型望远镜控制系统中的故障处理系统的设计

发布时间：2018-03-06 10:20

本文选题：故障处理　切入点：框架　出处：《中国科学技术大学》2014年硕士论文　论文类型：学位论文

【摘要】：为了满足我们对深空知识的渴求,现代天文望远镜的口径越来越大,硬件结构日趋复杂,安置地点也从地面向极地甚至太空延生。而这种发展趋势也使望远镜的维护困难,运行成本提高。为了方便工程人员维护,提高望远镜运行效率,降低观测成本,本文设计了针对大型望远镜控制系统的故障处理系统。该系统可用于帮助工程人员定位望远镜故障原因,获取关于故障的全面信息,提供故障解决辅助建议,甚至在某种程度上预警故障的出现。本文首先分析了现代天文望远镜的发展趋势,总结故障处理系统提出的背景并分析其对于大型望远镜控制系统的必要性。随后基于面向对象的程序设计与软件框架的方法对故障处理系统进行整体设计,并提出设计中所需考虑的几个问题。随后详细分析了故障处理系统中各个模块的详细设计,包括分布式日志服务、智能诊断专家系统和故障状态3D显示模块。其中智能诊断专家系统从整体设计出发,分别详细说明了故障树分析法和神经网络方法,以及基于这两种方法的智能诊断模块的设计和处理流程,同时针对这两种方法提出了改进分析。并通过对两种智能诊断专家系统的仿真数据测试,从测试的角度证明这两种方法用于故障诊断处理的可行性。最后总结了本文的创新性,包括： (1)以面向对象的软件框架方法设计了故障处理系统,满足可重用、模块化、可扩张等要求。 (2)设计了一个满足分布式需求的日志服务,不仅可在故障处理系统中使用,还可以用于所有需要日志服务的分布式系统中。 (3)设计了一个可使用故障树方法和神经网络方法两种不同方法的智能诊断专家系统,并通过测试验证其可行性。 (4)设计了基于OpenGL的故障状态3D显示模块,该模块不仅可用于故障处理系统,也可作为别的大型控制系统提供了一种新的信息交互手段。同时总结了未来可从向更泛化的软件框架、建立更精确的故障树模型和神经网络拓扑结构、寻找更多故障诊断方法等角度进行更深入的研究。
[Abstract]:In order to satisfy our thirst for knowledge of deep space, the modern astronomical telescope aperture is more and more big, the hardware structure of the growing complexity of resettlement sites from the ground to the polar or space extension. This trend also makes the telescope maintenance, the increase of the running cost. In order to facilitate the engineering staff to maintain, improve the operation efficiency of reducing the observation telescope. The cost of this paper designed for fault processing system of large telescope control system. This system can be used to help the designer to locate telescope fault reason, obtaining comprehensive information about fault fault, provide advice to solve, even warning fault to a certain extent.
This paper first analyzes the development trend of the modern astronomical telescope, it sums up the fault management system background and analyzes the necessity for large telescope control system. Then the method of program design and object-oriented software framework based on the fault diagnosis system for the overall design, the paper puts forward some problems that need to be considered in the design.
Then the detailed design of each module in the fault handling system is analyzed in detail, including distributed log service, intelligent diagnosis expert system and fault state 3D display module.
The intelligent diagnosis expert system starting from the overall design, each detailing the fault tree analysis method and neural network method, and based on the intelligent diagnosis module of the two methods of design and process, at the same time for the two methods are proposed to improve the analysis. And through the simulation data of two kinds of intelligent diagnosis expert system test. From the perspective of the test proved that the two methods for the feasibility of fault diagnosis.
Finally, the innovation of this article is summed up, including:
(1) the fault processing system is designed with the object-oriented software framework method, which satisfies the requirements of reusable, modularized and expansible.
(2) designed a log service that satisfies distributed demand. It can be used not only in fault handling system, but also in all distributed systems that need log service.
(3) an intelligent diagnosis expert system with two different methods, which can use fault tree method and neural network method, is designed and tested to verify its feasibility.
(4) a fault state 3D display module based on OpenGL is designed. It can be used not only in fault handling system, but also in other large-scale control systems.
At the same time, it summarizes the further research from the perspective of more generalized software framework, establishing more accurate fault tree model and neural network topology, and finding more fault diagnosis methods.

【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TH751;TP273

【相似文献】