摘要
由于电力负荷的不确定性,负荷模型难以建立,所以很难提供有效数据进行负荷管理。目前,利用基于环境信号的负荷模型参数辨识方法,虽然可以很频繁地进行负荷模型参数辨识,但由于数据的不确定性,造成产生很多不同识别结果。针对数据平台的这一特性,文中提出了一种负荷模型参数聚类方法,从辨识结果中提取具有代表性的负荷模型参数,为负荷管理的数据平台架构提供有效支撑。为了得到更好的聚类结果,采用基于故障后响应曲线的模型距离进行聚类。采用K-medoids聚类算法,簇数由簇半径决定。仿真结果表明了所提出的负荷模型参数聚类方法的有效性。
Because of the uncertainty of power load,it is difficult to establish load model,so it is difficult to provide effective data for load management.At present,load model parameter identification method based on environmental signals can be used frequently,but because of the uncertainty of data,many different identification results are produced.In view of this characteristic of the data platform,a load model parameter clustering method is proposed,by which representative load model parameters are extracted from the identification results,and effective support is provided for the data platform architecture of load management.In order to get better clustering results,the model distance based on post-fault response curve is used to cluster.K-medoids clustering algorithm is adopted,and the number of clusters is determined by the radius of clusters.The simulation results show the effectiveness of the proposed load model parameter clustering method.
作者
郑建锋
艾鸿宇
童志明
ZHENG Jian-feng;AI Hong-yu;TONG Zhi-ming(Quzhou Power Supply Company,Zhejiang Electric Power Corporation,Quzhou 324000,Zhejiang Province,China)
出处
《信息技术》
2020年第3期63-67,共5页
Information Technology
关键词
智慧电网
负荷管理
聚类算法
负荷模型
smart grid
load management
clustering algorithm
load model
作者简介
郑建锋(1977-),男,从事电网信息化建设工作。
