基于二次聚类的充电桩执行电价异常检测方法  

A method for detecting abnormal electricity prices at charging stations based on two rounds of clustering

作　　者：薛晓慧 张文[2] 张静 陈雁[2] 周春 陈亮[2] 曹增伟 XUE Xiaohui;ZHANG Wen;ZHANG Jing;CHEN Yan;ZHOU Chun;CHEN Liang;CAO Zengwei(State Grid Qinghai Electric Power Company,Xining 810008,China;Beijing China-Power Information Technology Co.,Ltd.,Beijing 100192,China)

机构地区：[1]国网青海省电力公司,青海西宁810008 [2]北京中电普华信息技术有限公司,北京100192

出　　处：《电信科学》2025年第1期184-190,共7页Telecommunications Science

基　　金：国家自然科学基金资助项目(No.72071070)。

摘　　要：由于电价政策复杂,执行环节多,监管难度大,电价执行错误现象时有发生,这不仅损害电力市场的公平性和效率,也影响电力企业的经济效益和用户的用电成本。提出了一种基于二次聚类的充电桩执行电价异常检测方法,首先进行电价执行异常分类及用电特征分析,其次通过K-means聚类算法剥离出电瓶车用户,进而在第二次聚类中采用含噪声应用的基于密度的空间聚类(density-based spatial clustering of applications with noise,DBSCAN)算法精确识别高价低接等更为复杂的违约情况。所提方法通过两次聚类分析,提高电价执行的准确性和效率,具有一定的理论意义和应用价值。Due to the complexity of electricity pricing policies,multiple implementation steps,and difficulty in regula‐tion,errors in electricity pricing implementation occur from time to time.This not only damages the fairness and effi‐ciency of the electricity market but also affects the economic benefits of power enterprises and the electricity costs of users.A method was proposed for detecting electricity price anomalies at charging stations based on secondary cluster‐ing.Firstly,the electricity price anomalies were classified and the electricity consumption characteristics were ana‐lyzed.Secondly,the K-means clustering algorithm was used to extract electric vehicle users.Then,in the second clus‐tering,the density-based spatial clustering of applications with noise(DBSCAN)algorithm was used to accurately identify more complex default situations,such as high price and low connection.The proposed method improves the accuracy and efficiency of electricity price implementation through two rounds of cluster analysis,and has certain theoretical significance and application value.

关 键 词：充电桩 执行电价异常 聚类分析 离群点检测 高价低接 

分 类 号：TM76[电气工程—电力系统及自动化]