基于FRFT和RBF神经网络的升船机配电网单相接地故障研究  

Research on single-phase grounding fault in distribution network of ship lift based on FRFT and RBF neural network

作　　者：张治臣 徐合力[1] 高岚[1] ZHANG Zhichen;XU Heli;GAO Lan

机构地区：[1]武汉理工大学船海与能源动力工程学院,湖北武汉430063

出　　处：《中国修船》2023年第6期48-52,共5页China Shiprepair

摘　　要：单相接地故障在陆地电网中为最常见的故障,若不能及时排查可能发展成相间短路等故障,严重威胁系统的安全性。文章引入馈线终端设备(FTU)对配电网进行监测,提出了一种基于分数阶傅里叶变换(FRFT)和径向基(RBF)神经网络结合的方法,针对单相接地故障进行识别。通过FTU监测配电网的电压数据,选取FRFT提取故障波形的能量率作为特征向量,最终利用训练过的RBF神经网络对故障进行识别定位。仿真试验结果表明:当FTU正常运行或发生信号丢失时,RBF神经网络均能够准确、高效地定位单相接地故障。Single-phase grounding fault is the most common fault in the land power grid.If not timely trouble-shot,it may develop into an inter-phase short circuit and other faults,seriously threatening the security of the sys-tem.This paper introduces feeder terminal unit(FTU)to monitor the distribution network,proposes a method based on fractional Fourier transform(FRFT)and radial basis function(RBF)neural network to identify single-phase grounding faults.It monitors the voltage data of the distribution network by FTU,and extracts the energy rate of the fault waveform by FRFT as the feature vector.Finally,the paper uses a trained RBF neural network to identify and lo-cate faults.The simulation results show that the RBF neural network can locate the single-phase grounding fault ac-curately and efficiently when the FTU runs normally,or the signal is lost.

关 键 词：配电网 故障区域定位 分数阶傅里叶变换 径向基神经网络 

分 类 号：U672[交通运输工程—船舶及航道工程]