摘要
为解决近年来用户行业变化特性加剧导致的难以准确辨识用户档案信息变动的问题,文中提出一种基于数据驱动的负荷特征异常辨识方法。首先,提出一种两阶段行业典型负荷形态构建方法,利用基于层次密度的含噪声应用空间聚类(hierarchical density-based spatial clustering of applications with noise,HDBSCAN)提取用户在不同场景下的典型日负荷曲线,并利用改进的K-means算法对提取出的典型日负荷曲线进行聚类分析,构建行业的典型负荷形态;其次,提出一种多维场景负荷特征异常智能研判方法,通过构造用户的负荷特征,使用熵权法评估行业典型场景的相对重要性,并采用单分类支持向量机(one-class support vector machine,OCSVM)算法量化每个场景下的用户负荷特征的异常程度,通过加权计算得到用户的综合嫌疑得分并排序,从而实现对负荷特征异常用户的准确辨识。最后,采用某地区实际用户数据进行算例验证。仿真结果表明,所提方法在行业典型负荷场景构建及负荷特征异常辨识方面表现出良好的可行性与实用价值。
To address the challenge faced by power grid companies in accurately detecting changes in user industryinformation,which has been complicated by the increasing variability of industry characteristics in recent years,a data-drivenapproach for identifying anomalies in load characteristics is proposed.Initially,a two-stage methodology for developing typicalload patterns for various industries is presented.The hierarchical density-based spatial clustering of applications with noise(HDBSCAN)technique is utilized to extract typical daily load curves for users under different scenarios.Subsequently,theseextracted daily load curves are clustered using an improved K-means algorithm to establish typical load patterns for therespective industries.In the second phase,a multidimensional intelligent diagnostic method for load characteristic anomalies isintroduced.User load characteristics are constructed,and the entropy weight method is employed to evaluate the relativesignificance of typical industry scenarios.The one-class support vector machine(OCSVM)algorithm is then utilized toquantify the degree of anomaly present in user load characteristics across each scenario.Comprehensive suspicion scores arecalculated and ranked to accurately identify users exhibiting abnormal load characteristics.The effectiveness of the proposedmethod is validated through the analysis of actual user data from a specific region.The results demonstrate that the method isboth feasible and practical for constructing typical industry load scenarios and for the identification of load characteristicanomalies.
作者
陈光宇
杨光
施蔚锦
蔡鑫灿
陈婉清
刘昊
CHEN Guangyu;YANG Guang;SHI Weijin;CAI Xincan;CHEN Wanqing;LIU Hao(School of Electric Power Engineering,Nanjing Institute of Technology,Nanjing 211167,China;Quanzhou PowerSupply Company of State Grid Fujian Electric Power Co.,Ltd.,Quanzhou 362000,China)
出处
《电力工程技术》
北大核心
2026年第2期70-79,共10页
Electric Power Engineering Technology
基金
国家自然科学基金资助项目(52107098)。
