配电网线路故障在线监测系统的设计与实现  

Design and implementation of online fault monitoring system for distribution network lines

作　　者：梁建权 张健 孙巍 赵雷雷 张德文 王悦 LIANG Jian-quan;ZHANG Jian;SUN Wei;ZHAO Lei-lei;ZHANG De-wen;WANG Yue(Electric Power Research Institute,State Grid Heilongjiang Electric Power Co.,Ltd.,Harbin 150001,China)

机构地区：[1]国网黑龙江省电力有限公司电力科学研究院,哈尔滨150001

出　　处：《信息技术》2022年第8期115-119,125,共6页Information Technology

摘　　要：利用Python开发的程序透过动态连接方式分别利用Open DSS所建立的配电网模型进行电力潮流解析,用PSS/E进行输电网计算来建构配电网故障在线实时监测分析系统。模拟实验结果显示,提出的两种算法均具有良好的分类性能和适应性,Mallat触发算法分类种类数目多且准确率高,抗干扰能力强,不论是在频率出现变动或是纯正弦波形、谐波失真、闪烁调变等状况,即使在外界信号干扰的情况下,检测结果的误差范围都在0~0.0015%之间。FINET+PSF的深度神经网路算法无需故障触发就能即时准确监测电弧故障,这使得实时监测配电网状态系统更加高效。The power flow is analysed through Python programs which utilizes distribution network model built by Open DSS through dynamic connection.And the PSS/E calculates the transmission network to construct an online system for real-time monitoring and analysis of distribution network faults.These two algorithms show both adaptability and classification.The Mallat trigger algorithm has various classification types,high accuracy,and strong anti-interference ability,whether under changing frequency,pure sine waveform,harmonic distortion,flicker modulation or other conditions.The error is just between 0 and 0.0015%,even being interfered by external signal.The deep neural network algorithm of FINET+PSF can detect arc fault timely and accurately without being triggered,making the system more efficient.

关 键 词：配电网故障监测 深度神经网络 Mallat触发算法 Python程序开发 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]