【作者】 闫岳婷；

【导师】 贾宏杰；

【作者基本信息】 天津大学 ， 电力系统及其自动化， 2009， 硕士

【摘要】 智能配电网的研究和实践正处于起步阶段,涉及众多技术领域,其中准确的状态估计是智能配电网研究的基础。大规模复杂电力网络的状态估计可采用分布式计算实现有效的求解。而配电网的可观性分析是状态估计的基础,因此深入探讨科学的适用于配电网状态估计分布式求解的可观测性分析方法,具有重要的理论意义和实用价值。本文主要工作如下:1.文中给出了一种有效的配电网络分割方法,能够将大规模的配电网络分割成给定数目的连通子网络,保证对各子网络实施状态估计时的计算量(本文称计算负担)近似相等。该分割方法由两部分构成:一是形成给定电力网络的深度优先搜索树;二是将网络分割成连通且计算负担基本相等的子网络的算法。利用IEEE-13节点等系统的仿真计算表明,该方法可以有效地实现大规模配电网络的分割,为实现配电网状态估计的分布式计算提供支持。2.研究了配电网络可观性分析算法:在分析量测岛规律基础上,根据电路基尔霍夫定理,推出每个量测岛内(注入量测已知)只需确定一个复电压变量即可实现状态估计的特点。进一步,利用可用注入量测对各量测岛进行合并。采取这一方法可以得到“唯一确定”的系统最大可观测岛,从而为分布式状态计算中的系统可观测性分析提供了可能。采用IEEE-13、IEEE-34和IEEE-37节点等系统验证了本文方法的正确性。本文工作对于寻求配电网络分割方法,以及在此基础上的网络可观测分析具有一定的帮助。更多还原

【Abstract】 Research on smart grid has been paid more and more attention nowadays. State estimation is one of the most concerned topics in this field, which can be effectively solved based on distributed computing technique. So, it is very important to find suitable methods to determine the system observability in the distributed state estimation. Work of this thesis is as following:1. An efficient method is presented to partition a large-scale distribution system into some connected sun-networks. And sizes of all sun-network are almost equal so that the computing burdens of the state estimation on them are also almost equal. The method consists of two steps: One is to form a weighted depth-first-search tree for the studied network. The other is to partition the system into some connected, balanced sub-networks. The presented method is validated by IEEE 13-bus system. It can be revealed that the method corrected and effective.2. Observability used for the distributed state estimation of distribution system is further studied. Based on Kirchhoff’s law in circuit theory, we find that a sub-network can be observed when an arbitrary node voltage is given and all its node injects are known. Also, some nodes in the sub-network can be combined in order to reduce the computing burden. The presented method can be used to determine the maximum observable island, which is useful for the distributed state estimation of distribution system. The presented method is finally validated by IEEE-13, IEEE-34 and IEEE-37 node systems. It is revealed that the presented method is very effective.Studies in this thesis are helpful to find more appropriate and effective state estimation method for distribution systems.更多还原