大规模膜法海水淡化系统故障诊断方法的研究

发布时间：2018-12-13 23:49

【摘要】：随着人口膨胀、工业化和城市化进程的不断加快以及人们生活水平的改善,水资源短缺问题备受瞩目,解决水资源短缺的海水淡化技术也引起了国内外学者和政府部门、国际组织的关注。在众多海水淡化方法中,反渗透法因其低能耗、高效率、高产水品质等优点得到广泛的应用。反渗透系统的运行受到海水水质和核心组件反渗透膜性能等多方面的影响。为了保证大型海水淡化工程的正常运行,避免造成重大的经济损失和人员损失,对大规模海水淡化系统的故障诊断至关重要。由于大规模膜法海水淡化系统设备复杂、运行过程存在不稳定因素以及反映系统运行状态的传感器等设备不精确等问题,单一故障诊断方法无法完整反映系统故障状态,不足以解决大规模膜法海水淡化系统故障诊断问题。针对舟山六横岛日产十万吨级膜法海水淡化系统的反渗透部分故障问题,以膜组件和能量回收装置为主要研究对象,系统分析了影响反渗透海水淡化系统运行状态的各方面因素。在深入分析能量回收装置故障机理的基础上,构建了以“能量回收装置故障”为顶事件的故障树。利用二元决策图方法对引用的反渗透系统故障树进行定性和定量分析,根据改进的组件连接法连接规则,得到相应的二元决策图;采用基于哈夫曼码的割集排序方法,通过比较哈弗曼码长度,从而快速有效地确定需要检测故障源的排序。详细介绍了D-S证据理论的基本概念、合成规则和D-S证据推理信息融合决策的基本过程,建立了基于D-S证据理论的反渗透海水淡化系统故障诊断模型,提出了符合反渗透海水淡化系统诊断特点的基本概率分配方法,构造了符合系统诊断特点的决策规则,通过实例验证了多证据体故障特征利用信息融合故障诊断技术的可行性和可信度。克服了单一故障诊断方法的缺陷与不足,提高了系统的有效性和鲁棒性。基于.NET的C#编程语言,利用其面向对象的继承性和可移植性等优点,结合SQL Server 2005数据库,设计并开发了一套基于D-S证据理论的信息融合计算分析软件,介绍了软件计算分析流程和主要功能,提高了计算效率,减少了人工计算的劳动力和出错率。
[Abstract]:With the expansion of population, the accelerating process of industrialization and urbanization, and the improvement of people's living standard, the shortage of water resources has attracted much attention. The desalination technology to solve the shortage of water resources has also caused scholars and government departments at home and abroad. Concerns of international organizations Among many desalination methods, reverse osmosis has been widely used because of its advantages of low energy consumption, high efficiency and high water quality. The operation of reverse osmosis system is affected by seawater quality and the performance of core components of reverse osmosis membrane. In order to ensure the normal operation of large-scale seawater desalination project and avoid causing heavy economic loss and personnel loss, it is very important for the fault diagnosis of large-scale seawater desalination system. Because of the complex equipment of large-scale membrane seawater desalination system, the unstable factors in the operation process and the imprecision of the equipment such as sensors which reflect the operating state of the system, the single fault diagnosis method can not completely reflect the fault state of the system. It is not enough to solve the problem of fault diagnosis of large-scale membrane desalination system. Aiming at the partial failure of reverse osmosis (RO) in the 100,000-ton membrane desalination system in Liuheng Island, Zhoushan, the main research object is membrane module and energy recovery device. The factors affecting the operation state of reverse osmosis seawater desalination system are analyzed systematically. On the basis of deeply analyzing the fault mechanism of energy recovery device, a fault tree with "energy recovery device fault" as its top event is constructed. The fault tree of the referenced reverse osmosis system is qualitatively and quantitatively analyzed by using the method of binary decision graph, and the corresponding binary decision graph is obtained according to the connection rules of the improved component connection method. By comparing the length of Haverman code, the sorting method of cut set based on Huffman code is used to determine the sort of fault source quickly and effectively. This paper introduces the basic concept of D-S evidence theory, the synthesis rules and the decision process of D-S evidence inference information fusion, and establishes the fault diagnosis model of reverse osmosis seawater desalination system based on D-S evidence theory. The basic probability distribution method which accords with the diagnostic characteristics of reverse osmosis seawater desalination system is put forward, and the decision rules that accord with the diagnostic characteristics of the system are constructed. The feasibility and reliability of fault diagnosis technology based on information fusion are verified by examples. The shortcomings and shortcomings of single fault diagnosis method are overcome, and the effectiveness and robustness of the system are improved. Based on the C # programming language of .NET, this paper designs and develops a set of information fusion computing and analysis software based on D-S evidence theory, using the advantages of object oriented inheritance and portability, combined with SQL Server 2005 database. The flow and main functions of software calculation and analysis are introduced. The efficiency of calculation is improved and the labor force and error rate of manual calculation are reduced.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P747

【相似文献】