基于多维特征参量的台区拓扑识别方法研究  

Research on a topology identification method for distribution areas using multi-dimensional electrical characteristic parameters

作　　者：王珺 朱亮 高晶 康守信 章杨帆 游小辉 伍栋文 WANG Jun;ZHU Liang;GAO Jing;KANG Shouxin;ZHANG Yangfan;YOU Xiaohui;WU Dongwen(State Grid Jiangxi Power Supply Service Management Center,Nanchang 330032,China;Beijing Smart Chip Microelectronic Technology Co.,Ltd.,Beijing 102200,China;State Grid Nanchang Electric Power Supply Company,Nanchang 330095,China)

机构地区：[1]国网江西省电力有限公司供电服务管理中心,南昌330032 [2]北京智芯微电子科技有限公司,北京102200 [3]国网江西省电力有限公司南昌供电分公司,南昌330095

出　　处：《浙江电力》2024年第12期114-121,共8页Zhejiang Electric Power

基　　金：国家电网有限公司科技项目(52180024000G);国网江西省电力有限公司科技项目(52185223000L)。

摘　　要：针对传统的单一电气特征参量拓扑识别方法准确率和稳定性不高的问题,提出基于多维电气特征参量的台区拓扑识别方法。首先,通过信标广播绝对时间与信标传输时延补偿完成台区内节点时间同步;其次,结合节点数据同步采集和边缘处理,建立高效的数据压缩机制,实现多维特征参数的高效采集和传输;然后,通过监听周围节点创建节点的邻节集合,减小拓扑计算的范围;最后,基于电压、电流和电量特征参量,提出基于模糊矩阵算法的多维特征参量拓扑识别方法,并结合算例进行验证。结果表明,所提出的多维特征参量拓扑识别方法准确率高于94.12%,准确率和鲁棒性均优于基于传统的单一特征参量拓扑识别方法。To address the issues of low accuracy and stability in traditional topology recognition methods that rely on single electrical characteristic parameters,a new method for topology identification in distribution areas using multi-dimensional electrical characteristic parameters is proposed.Firstly,synchronization of node time within distribution areas is achieved by compensating for the absolute time broadcast by the beacon and the transmission delay of the beacon.Secondly,by using synchronous node data acquisition and edge processing,an efficient data compression mechanism is established to enable the efficient collection and transmission of multi-dimensional characteristic parameters.Additionally,by monitoring surrounding nodes,a neighboring node set is created to reduce the scope of topology calculations.Finally,the paper proposes a topology identification method based on the fuzzy matrix algorithm,utilizing voltage,current,and power characteristic parameters,and validates it with examples.The results demonstrate that the proposed method achieves an accuracy of over 94.12%,surpassing traditional topology recognition methods that rely on single electrical characteristic parameters in both accuracy and robustness.

关 键 词：配电台区 拓扑识别 同步采集 邻节集合 模糊矩阵 

分 类 号：TM73[电气工程—电力系统及自动化]