摘要
为了实现准确的反窃电识别定位,提出基于大数据高维随机矩阵的反窃电识别定位方法。构建反窃电定位的节点分布模型,采用大数据挖掘的方法及逆行反窃电定位识别的随机分布结构重组,进行反窃电定位过程中的随机分布和信息融合。以电网的高频分量差异性特征量作为统计特征量,进行直流断路控制设计,基于大数据高维随机矩阵重构的方法,进行反窃电识别定位过程中的特征空间重构,采用动态调节的方法进行反窃电识别定位的动态分布融合,建立反窃电识别定位的有功功率分析模型,通过负荷的状态差异性,实现反窃电识别定位。仿真结果表明,采用该方法进行反窃电识别定位的精度较高,参数估计准确性较好,提高了反窃电识别定位诊断分析能力。
In order to improve the ability of intelligent alarm and diagnosis of power dispatching faults,a design method of intelligent alarm system for power dispatching faults based on data driving is proposed,and an intelligent alarm information collection model for power dispatching faults is established.Data-driven control of intelligent alarm process for power dispatching,and cross-compilation technology is combined to extract fault features and load instructions for power dispatching faults.The fuzzy detection method is used to fuse and judge the intelligent fault information of power dispatching faults,and the detection statistics and decision functions are constructed.The data-driven and big data information clustering methods are used to realize the feature fusion and classification recognition of power dispatching faults,and the intelligent alarm and information fusion model of power dispatching faults is established.The multi-information fusion and feature ambiguity detection methods are used to realize the intelligent alarm and judgment of power dispatching faults and improve the power dispatching fault detection and recognition ability.Simulation results show that the intelligent diagnosis of power dispatching fault detection by this method is better,and the ability of fault detection and identification is improved.
作者
薛峪峰
张俊超
马晓琴
XUE Yufeng;ZHANG Junchao;MA Xiaoqin(State Grid Qinghai Information&Telecommunication Company State Grid Qinghai Date Center,Xining 810008,China)
出处
《自动化与仪器仪表》
2020年第11期220-222,226,共4页
Automation & Instrumentation
基金
青海省科学技术项目
基于大数据智能反窃电管理研究(No.SGQHXTOOGCJS2000146)。
关键词
大数据
高维随机矩阵
反窃电
识别
差异性特征
data-driven
electric power dispatching
malfunction
intelligent alarm
keywords information clustering,intelligent diagnosis
作者简介
薛峪峰(1984-),男,陕西韩城人,工程师,主要研究方向为电力营销信息化及其应用。
