基于图论的矩阵算法在配电网故障定位中的应用

分析现有配电网故障定位算法的不足，提出一种基于图论知识的改进配电网故障定位算法。该算法首先利用配电网有向拓扑结构生成一个可反映配电网拓扑信息的网络拓扑矩阵，当配电网发生故障时，安装在开关处的馈线终端（FTU）通过通用分组服务技术（GPRS）网络向主站上传故障信号，生成故障信息矩阵，进而得到故障判定矩阵，由故障判据定位故障发生区段。该算法不仅可以对单电源的配电网单重故障定位，对于多电源网络的多重故障以及馈线末端故障也可作出快速判断，同时对于FTU上传的不完备故障信息情况也同样适用。算例测试结果表明，该算法判断简洁直观，计算量小，实用性强，满足现代配网自动化要求。

Fault Location of Distribution Network by Applying Matrix Algorithm Based on Graph Theory

Based on an analysis of the existing weakness of fault location algorithm for distribution network, an improved fault location algorithm based on graph theory is proposed in this paper. Firstly a network topology matrix is produced according to the distribution network. When a fault occurs, the fault information matrix can be generated from the fault signal uploaded from FTU through GPRS. Then a fault judgement matrix is obtained by adding the two matrixes above, and the fault section can be located according to the failure criterion. The proposed algorithm can not only locate a single fault of the distribution network with single power source, but also make a quick judgement of the multiple faults and the feeder end fault of the distribution network with multiple power sources. Furthermore, the algorithm is also applicable to the incomplete fault information from FTU. The simulation results show that the proposed algorithm is concise and practical with low computation cost, and can satisfy the requirements of modern distribution network automation.