基于模体的低压配电网负荷波动特性分析  

Motif-based Analysis on Load Fluctuation Characteristics in Low-voltage Distribution Network

作　　者：李知艺 马翔宇 于群 辛焕海[1] 鞠平 LI Zhiyi;MA Xiangyu;YU Qun;XIN Huanhai;JU Ping(College of Electrical Engineering,Zhejiang University,Hangzhou 310027,China;Polytechnic Institute,Zhejiang University,Hangzhou 310015,China;College of Energy and Electrical Engineering,Hohai University,Nanjing 211100,China)

机构地区：[1]浙江大学电气工程学院,浙江省杭州市310027 [2]浙江大学工程师学院,浙江省杭州市310015 [3]河海大学能源与电气学院,江苏省南京市211100

出　　处：《电力系统自动化》2022年第10期209-215,共7页Automation of Electric Power Systems

基　　金：国家自然科学基金资助项目(U2066601,52007164)。

摘　　要：随着新型电力系统的建设和发展,低压配电网中负荷内涵日渐丰富,负荷波动愈加随机,亟须挖掘用户侧海量量测数据的隐藏价值。为快速、精准辨识负荷演变态势,借鉴模体理念,提出了一种数据驱动、简单高效的低压配电网负荷波动特性分析新思路。针对单个负荷,将连续时间点负荷功率的相对大小作为准则构建个体模体,提取负荷连续波动的细节特征,进而识别负荷类型;针对负荷集群,将连续时间段负荷功率整体波动的相对距离作为准则构建集群模体,刻画分散负荷的聚合波动特征,进而推断外界因素影响下的集群波动规律。基于多个实际数据集的数值实验表明,所提方法能从个体、集群2个层面实现对低压配电网负荷波动特性的准确认知。With the construction and development of new power systems,the load connotation of low-voltage distribution networks is becoming increasingly extensive,and the load fluctuations are more irregular.It is essential to fully mine the hidden value of the massive measurement data on the demand side.To quickly and accurately identify the load evolution situation,a novel idea for a data-driven,simple and efficient analysis of the load fluctuation characteristics of the low-voltage distribution network is proposed based on the concept of the motif.For the individual loads,the relative relation of load power in a continuous time period is used as the criterion for the individual motif construction,and the detailed characteristics of load variations are extracted to identify the attribute of load;for the load clusters,the relative distance of overall load power fluctuations in a continuous-time period is adopted to construct clustered motifs,which can describe the aggregated fluctuation characteristics of the dispersed load,and infer the fluctuation rules of load clusters under the influence of external factors.Numerical experiments based on the multiple actual datasets show that the proposed method can achieve the accurate cognition of load fluctuation characteristics of low-voltage distribution networks at the individual and cluster levels,respectively.

关 键 词：配电网 电力负荷 大数据 时间序列 模体 波动特性 

分 类 号：TM714[电气工程—电力系统及自动化]