【作者】 谢国财；

【导师】 李朝晖； 毕亚雄；

【作者基本信息】 华中科技大学 ， 水利水电工程， 2012， 博士

【摘要】 水电是中国当前最主要的清洁能源,在全世界日益重视节能减排的背景下,如何在确保水电机组安全稳定运行的前提下尽可能提高其利用率便成为十分突出的研究课题。自上世纪90年代以来,旨在提高水电机组运行可靠性和经济效益的“状态检修”成为研究的焦点,以掌握水电机组健康状况为目标的状态监测技术取得了长足的发展与进步,多种水电机组状态监测系统投入工业应用。这些系统基本都是专项监测装置或集合监测系统,在保证水电机组的安全稳定运行、提高设备利用率等方面发挥了一定作用。然而,这些系统普遍没有实现机组的全面监测,也没能将各专项监测集成在同一平台上,因而专项监测装置之间难以做到信息共享与行为协同(如同步采集、同步存储、关联分析等)。加之,水电机组是一个由水力、机械、电气、控制和辅助设备构成的强耦合复杂大系统,往往牵一发而动全身。当前的各种监测方法不能有效解决多来源、多时标、异构机组状态数据的有机融合问题,从而不能实现机组的综合分析与诊断。为此,提出了基于团队智能的水电机组集成监测方法,其基本目标是：运用团队智能技术,将各监测装置通过实时现场总线网集成到同一平台上进而构成一个有机的机组集成监测团队,建立各监测装置之间关联协同的机制,共同实现多来源、多时标及异构机组状态数据的有机融合,便于进行机组的综合分析与诊断。通过总结团队的鲜明特征,引出了团队智能(TI)的概念并给出了其定义,并对团队中可能出现的团队行为进行了形式化描述。系统地研究了团队模型,包括组织结构、任务描述、角色关系、组织规范及人事关系等内容；通过任务描述,实现了团队目标、行为与任务的统一,并可将团队的协同行为归结为组织规范约束下的角色关系。建立了基于消息机制的三层垂直分布结构团队成员模型(TIP),并分别对成员模型的消息处理层、行为控制层及核心功能层的框架进行了设计。根据团队智能理论,建立了水电机组集成监测团队框架；整个监测团队由控制单元TIP、机械单元TIP、电气单元TIP、综合单元TIP及其它TIP组成,并分别对这些TIP的核心功能进行了阐述；讨论了机组集成监测团队的基本运行机制,包括分工、协同、制约及冲突等。通过时间融合、运行工况融合及异常事件融合等关键技术实现机组状态数据的有机融合。围绕机组海量状态数据的组织、存储与管理等问题,分别研究了机组状态数据的优化组织策略、选择性智能存储策略以及有效性验证与管理。在实现机组集成监测的基础上,探讨了水电机组的综合分析方法。将传统的阈值分析与机组运行工况、异常事件相结合,研究了基于动态阈值的实时故障预警方法,并介绍了该方法在变压器故障预警方面的应用实例。以机组集成监测提供的全面、关联性强的状态数据为基础,介绍了机组稳定性综合分析的现场案例。针对调速油系统常见的异常或故障,分别定义了能较好反映它们的性能指标并给出了相应的计算方法；以状态数据与性能指标为基础,探讨了调速油系统性能分析与异常检测的方法；最后给出了调速油系统实时监测、日分析、漏油检测及故障预警等应用实例。实践证明,基于团队智能的水电机组集成监测为开展设备的综合分析与诊断提供了优良的平台,这对提高分析与诊断结果的准确性、可靠性及可信性具有极其重要的作用。更多还原

【Abstract】 Hydroelectricity is the main clean energy at present in China. Under the background of the worldwide increasing attention to energy saving and emission reduction, It’s a prominent issue that how to improve the utilization of Hydro Turbine Generator Sets (HTGS) as much as possible when the save and stable operation of HTGS can be guaranteed. Since the 1990s, condition based maintenance aimed at improving the reliability of HTGS and economic benefits has become the focus of research. The technique of condition monitoring used to master the health condition of HTGS has made considerate development and progress, and a variety of condition monitoring systems of HTGS have been put into industrial application. These systems are basically special monitoring devices or collective monitoring systems, which play a role in ensuring the save and stable operation of HTGS and improving the facility utilization.However, these systems neither achieve overall monitoring generally, nor can make all special monitoring devices be integrated into the same platform, and thus information sharing and behavior collaborating (such as synchronous sampling, synchronous storage, correlation analysis) among special monitoring devices are difficult to be realized. Additionally, HTGS is a close-coupling, complex, and large system consisted of hydraulic, mechanical, electrical, control and auxiliary equipment, whose health condition can be affected by any one part. The current monitoring methods cannot solve organic fusion issue of state data of HTGS that has the characteristics of multi source, multi time scope, and isomerization, so the synthetic analysis and diagnosis of HTGS cannot be achieved.For this reason, the method for integrated monitoring of HTGS based on Team Intelligence (TI) is proposed. The basic objective of this method is to make all special monitoring devices be integrated into the same platform via real-time fieldbus and then form an organic integrated monitoring team of HTGS; the correlation and coordination mechanisms of these special monitoring devices are established; they jointly realize organic fusion of state data of HTGS that has the characteristics of multi source, multi time scope, and isomerization, and as a result it’s convenient for the synthetic analysis and diagnosis of HTGS.By summarizing the distinctive characteristics of teams, the concept of TI is drawn forth and its definition is given, and the formal descriptions of team behaviors possibly existed in a team are conducted. The research of team model is carried out systematically, including organizational structures, task descriptions, role relations, organizational norms and human relations, etc.. The unity of goals, behaviors and tasks of the team is achieved through task descriptions, and the collaborative behaviors can be contributed to the role relations under constraint of organizational norms in a team. The model of Team Intelligence Player (TIP) with three-layer vertical structure based on the message mechanism is established, and the frameworks of message processing layer, behavior control layer, and core functional layer included in the model of TIP are separately designed.According to TI theory, the framework of the integrated monitoring team of HTGS is established, which is composed of control unit TIP, mechanical unit TIP, electrical unit TIP, synthetic unit TIP and other TIPs, and core functions of these TIPs are discussed separately. The basic operating mechanisms of the integrated monitoring team of HTGS are elaborated, including divisions of work, coordination and cooperation, constrains and conflicts, etc.. Organic fusion of state data of HTGS is achieved by key techniques such as time fusion, operating condition fusion and abnormal event fusion, etc..Surrounded by the problems of organization, storage and management of mass state data of HTGS, optimizing organization strategy, selectively intelligent storage strategy, and validity confirmation and management of state data of HTGS are separately studied.Based on integrated monitoring of HTGS, the synthetic analysis approach to HTGS is discussed. Making the conventional threshold analysis combine with operating conditions and abnormal events of HTGS, the method for real-time fault prewarnings based on the dynamic threshold is studied, whose application examples in the fault prewarning of the transformer are introduced. On basis of the overall and strong relevance state data provided by integrated monitoring of HTGS, the field case of the synthetic analysis for the stability of HTGS is presented. Aimed at common failures or faults of the Oil System of the Governor (OSG), the performance indices that can well reflect them are defined separately, and the calculation methods are given. Based on state data and performance indices, the approaches to the performance analysis and failure detection of OSG are introduced. In the end, the application examples such as the real-time monitoring, daily analysis, oil leakage, and fault prewarning of OSG are given.It has been proved that the integrated monitoring of HTGS based on TI can provide a good platform for carrying out the synthetic analysis and diagnosis of facilities, which plays extremely important roles in improving the accuracy, reliability, and credibility of the analysis and diagnosis results.更多还原