基于边缘计算技术的变电站二次设备异常判定  

Anomaly Determination of Secondary Equipment in Substation Based on Edge Computing Technology

作　　者：张舸 吴奇珂 黎翔 刘奇 程培军 ZHANG Ge;WU Qike;LI Xiang;LIU Qi;CHENG Peijun(Power Dispatching and Controlling Center(System Operation Department),Guangzhou Power Supply Bureau of Guangdong Power Grid Co.,Ltd.,Guangzhou Guangdong 510620,China)

机构地区：[1]广东电网广州供电局电力调度控制中心(系统运行部),广东广州510620

出　　处：《电子器件》2025年第1期164-169,共6页Chinese Journal of Electron Devices

基　　金：南方电网科技项目(080016KK52200004)。

摘　　要：针对电网不断扩大规模下的变电站二次设备管理过程中设备状态判定困难及失误率高的问题,提出了基于边缘计算技术的变电站二次设备异常判定方法。分析了变电站内二次设备异常数据以及二次设备构建方法,提出了设备元件异常度计算方法,建立了故障告警分析模型。并且针对二次设备中数据分类问题提出了改进随机森林算法。提出了二次设备预警及故障诊断系统,结合实际案例进行了仿真分析,说明了系统的有效性,能够有效提升二次设备异常状态的判断准确率,相比现有系统提升3.2%。In response to the challenges of difficulty in equipment status determination and high error rate in secondary equipment man-agement in the context of the expanding scale of the power grid,a method for abnormal determination of secondary equipment in substa-tions based on edge computing technology is proposed.This method involves analyzing abnormal data from secondary equipment within the substation and the construction of secondary equipment.A method for calculating the abnormality degree of equipment components is presented,and a fault alarm analysis model is established.Furthermore,an improved random forest algorithm is introduced to address the data classification issue in secondary equipment.A secondary equipment early warning and fault diagnosis system is also proposed,and through simulation analysis using real-world cases,it is demonstrated that the system is effective.It can significantly improve the ac-curacy of abnormal status determination in secondary equipment,achieving a 3.2%increase compared to existing systems.

关 键 词：边缘计算 二次设备 异常判定 改进随机森林 

分 类 号：TM712[电气工程—电力系统及自动化] TP181[自动化与计算机技术—控制理论与控制工程]