【作者】 谭熙静；

【导师】 何正友；

【作者基本信息】 西南交通大学 ， 电工理论与新技术， 2012， 硕士

【摘要】 随着用电负荷的快速增长、供电可靠性要求的提高,使得以城市、负荷集中区域为代表的配电网规模逐渐增大、拓扑结构越来越复杂；加之陆续有分布式能源的接入、电动汽车充电站的应用,配电网的运行复杂性也大大增加。因此,传统的针对已发生故障进行处置的配电网故障诊断技术已经不能适应复杂配电网安全、可靠、经济运行的需要。所以,对复杂配电网的健康诊断进行研究讨论是非常有必要的。将复杂配电网当作一个生命体来研究,介绍了复杂配电网系统健康诊断和评估的概念,在将复杂配电网运行健康状况与人的健康状态相类比的基础上,使用健康指标体系描述复杂配电网的功能和性能表征。从主、次作用因子,内、外影响因素等方面对指标进行筛选,同时明确各项指标的健康标准,建立一套衡量复杂配电网健康状况的定量参考系,通过与参考系的比较,找出系统自身差距和变化态势,对己造成或可能发生的病症异常,及早给出响应动作建议,保障其处于稳定健康状态。对某地10kV配电网进行实例分析,从小时间尺度上系统的健康评估和大时间尺度上的健康趋势预测两个方面对该诊断方法的适用性和正确性进行验证。隐藏故障的研究是近年来电网故障研究的新热点,若互感器存在隐藏故障,将对电网故障时继电保护等设备的正常工作造成直接影响。现有研究多针对继电保护、断路器等设备进行,甚少涉及对互感器隐藏故障的在线识别。对互感器可能存在的隐藏故障类型进行分析总结,使用潮流追踪技术和解析冗余方法建立各支路功率和电流之间的关系式,通过验证其残差向量是否满足约束条件来对该互感器的隐藏故障进行识别。通过对一个5节点配电网网络的算例分析,验证了该方法对互感器存在的隐藏故障等异常具有较好的识别能力。给出了一种对复杂配电网进行健康诊断的思路,有助于调度人员及时掌握系统健康状况与趋势,实现在时间、空间和内容上对以往故障诊断的拓展。同时,对互感器隐藏故障的识别进行探索性研究,旨在对现有的隐藏故障研究进行补充和完善。更多还原

【Abstract】 With the rapid growth in electricity loads and the improvement of power supply reliability, the scale of the distribution grid represented by cities and load concentrated regions is getting bigger and bigger, and its topology is more and more complex. Its operational complexity is greatly increased because of the access of distributed energy and the application of the electric car charging station at the same time. Such, the traditional fault diagnosis technology aimed at occurred fault of distribution network cannot meet the demand of safe, reliable and economic operation. Therefore, it is quite necessary to discuss the real-time health diagnosis of complex distribution grid (CDG).Regarding CDG as a living body, introduce in the concept of CDG health diagnosis and use health index system to describe its functionality and performance conditions by means of the analogy of CDG health status and human health state.Screen index by considering primary-secondary effect as well as internal and external influence factors. Determine the health standard of each index and establish a quantitative reference frame to measure its health state. Find its gap and changing trend by compared with the reference frame, and then give response action suggestions to the abnormality occurred or likely to occur to guarantee that it is in a stable health state. Verify the applicability and accuracy of the diagnostic structure by analyzing certain real lOkV distribution network including health assessment of the system in small time scales and health trend prediction in large time scales.The study of hidden failure is a new focus of power system fault diagnosis in recent years. The hidden failure of CT/PT has a direct impact on the normal work of equipments such as relay protection etc. Few studies have been done on the research of transformer hidden failure in contrast with the study of relay protection and breaker etc. The main manifestations of transformer hidden failure is analyzed and summarized in this paper. Establish the relationship between every branch power and its current using power flow tracing technology and Analytical Redundancy Relation (ARR), and recognition its hidden failure by verifying whether the residual vector satisfy the constraint condition or not. The analysis result of a5nodes distribution network shows that this method is capable to recognize the hidden failure and other abnormality. This paper finally gives a method of CDG health diagnosis which is helpful for dispatchers to master the health state and operation trend of the system, and achieve the expansion of the previous fault diagnosis in content, time and space. Meanwhile, study on the recognition of the hidden failure of CT/PT, which could be a supplementary to the research.更多还原