摘要
针对单相接地故障频发且难以准确定位故障发生区段的问题,提出了一种基于数据-物理融合的单相接地故障两阶段诊断方法。首先,根据集中型馈线自动化系统的诊断流程,通过改进的极限学习机算法提高馈线自动化故障识别能力。然后,预测故障发生时刻各配变的负荷值,累加得到各开关的负荷占比,进而比较故障发生时馈线负荷骤降度与各开关负荷占比,对故障区段进行定位。最后,结合两阶段的诊断结果确定最终的故障发生位置并进行隔离。通过仿真验证,表明所提方法能够有效提高单相接地故障诊断的准确性。
To solve the problem of frequent single-phase grounding faults and difficulty in accurately locating the fault occurrence section,a method based on data-physics integration two-stage location of single phase grounding faults was proposed.Firstly,according to the research and judgment process of centralized feeder automation(FA),the improved Extreme Learning Machine(ELM)algorithm was used to improve the FA fault identification and analysis ability.Then,the load value of each distribution transformer at the time of fault occurrence was predicted,and the load proportion of each switch was obtained by accumulating.Subsequently,the load drop of the feeder and the proportion of the load of each switch were compared when the fault occurs,and the fault section was located.Finally,combined with the results of the two stages,the final fault location was determined and isolated.Through simulation verification,it is shown that the proposed method can effectively improve the accuracy of judging and isolating single-phase grounding faults.
作者
张小勇
李自乾
韩敏
夏旺
黄维成
ZHANG Xiaoyong;LI Ziqian;HAN Min;XIA Wang;HUANG Weicheng(Pingliang Electric Power Supply Company,State Grid Gansu Electric Power Company,Pingliang 744000,China)
出处
《机械与电子》
2025年第7期24-29,共6页
Machinery & Electronics
基金
国网甘肃省电力公司科技项目(52270923000A)。
关键词
单相接地故障
集中型FA
故障诊断
负荷预测
极限学习机
single-phase ground fault
centralized FA
fault judgment
load forecasting
extreme learning machine
作者简介
张小勇(1977-),男,甘肃平凉人,高级工程师,研究方向为配电网运行维护和管理;李自乾(1994-),男,甘肃平凉人,工程师,研究方向为电网运行、智能电网技术与设备。
