摘要
为了降低电网运行的技术损耗、预防用户窃电行为,文章从线损计算方法着手,分析用于窃电诊断的指标体系;基于Elman神经网络以及k-means聚类分析算法构建线损预测模型;基于高维随机矩阵构建窃电诊断模型,融合2个模型设计了电网线损与窃电预警平台。经过配电网实例数据预测与仿真,线损预测误差较小,窃电诊断定位准确,为降低损耗、预防窃电以及后续电网规划、策略优化提供了智能化分析手段。
In order to reduce the loss of power grid operation technology,prevent power thefts,this paper starts with the line loss calculation method,and uses it to the power diagnosis of index system based on Elman neural network and k-means cluster analysis algorithm to construct line loss prediction model.A power theft diagnosis model is established based on the higher dimensional random matrix.The integration of the two models yields a platform of grid line loss and power theft warning.Through the distribution network prediction and simulation instance data,the line loss prediction error is small,power diagnosis is accurate.It can reduce line loss,prevent power theft,and provides intelligent analysis means for strategy optimization.
作者
胡程平
赵扉
陈婧
胡剑地
林超
HU Chengping;ZHAO Fei;CHEN Jing;HU Jiandi;LIN Chao(State Grid Jiaxing Power Supply Company,Jiaxing 314001,China;SGIT-Great Power,Fuzhou 350003,China)
出处
《微型电脑应用》
2022年第11期91-94,共4页
Microcomputer Applications
关键词
聚类分析
ELMAN神经网络
线损预测
窃电诊断
cluster analysis
Elman neural network
line loss prediction
power theft diagnosis
