工业电力系统基于录波数据的故障元件智能诊断方法研究  

Research on Intelligent Diagnosis Method of Faulty Components in Industrial Power System Based on Recorded Wave Data

作　　者：张赛鹲 Zhang Saimeng(State Grid Xianyang Power Supply Company,Xianyang Shaanxi 712000,China)

机构地区：[1]国网咸阳供电公司,陕西咸阳712000

出　　处：《现代工业经济和信息化》2024年第6期145-147,共3页Modern Industrial Economy and Informationization

摘　　要：利用模糊C均值聚类算法从录波数据中提取各种工业电力系统的故障特征,例如电压、电流相量特征、小波系数比特征和故障分量特征等。通过聚类分析,计算输电元件的故障期望值,若计算结果超过阈值,则判定该元件存在故障。仿真模拟结果表明,此次建立的诊断方法在工业电力系统中能够有效检测出故障元件。Fuzzy C-mean clustering algorithm is used to extract various industrial power system fault characteristics from the recorded waveform data,such as voltage and current phase characteristics,wavelet coefficient ratio characteristics and fault component characteristics.Through the clustering analysis,the fault expectation value of the transmission component is calculated,and if the calculated result exceeds the threshold value,the component is determined to be faulty.Simulation results show that the diagnostic method established in this work can effectively detect faulty components in industrial power systems.

关 键 词：录波数据 电力系统 故障元件 智能诊断方法 

分 类 号：TM732[电气工程—电力系统及自动化]