基于神经网络的多电源配电网故障定位  

Fault Location of Multi-sources Power Distribution Network Based on Neural Network

作　　者：阮祥淼[1] 张馨心[2] 

机构地区：[1]信阳供电公司电力经济技术研究所,河南信阳464000 [2]平顶山天安煤业股份有限公司二矿,河南平顶山467000

出　　处：《煤矿机械》2014年第2期239-240,共2页Coal Mine Machinery

摘　　要：风光类分布式电源的接入使得配电网的故障定位问题由单电源模式变为多电源模式。为了研究多电源配电网的故障定位,研究了应用神经网络进行电源配电网故障定位的可行性。通过训练样本对神经网络进行训练,实现了故障过电流信息到故障区域的映射。算例的结果证明应用神经网络求解多电源配电网故障区域是可行的,具有一定的理论指导意义。The fault location problem of power distribution network changes from single to multilateral model. This paper studies the feasibility of application of neural network for power distribution network fault location to research the fault location of muff-sources power distribution network. The mapping from fault current information to faulty section was implemented by the neural network weight training guided by training sample. The results of example prove that it is feasible to find the faulty section in muti-sources power distribution network using neural network, and this method has a certain theoretical guiding significance.

关 键 词：多源模式 神经网络 故障定位 

分 类 号：TM76[电气工程—电力系统及自动化]