基于改进关联规则算法的用电行为险态自动化辨识模型  

An Automatic Identification Model of Power Consumption Risk Based on Improved Association Rule Algorithm

作　　者：岳恒 董俐君 邸金龙 YUE Heng;DONG Lijun;DI Jinlong(State Grid Huitongjincai(Beijing)Information Technology Co.,Ltd.,Beijing 100053,China)

机构地区：[1]国网汇通金财(北京)信息科技有限公司,北京100053

出　　处：《微型电脑应用》2025年第1期201-204,208,共5页Microcomputer Applications

摘　　要：基于改进关联规则算法提出一种用电行为险态自动化辨识模型,能精准挖掘异常用电数据,提升用电行为险态辨识效果。通过引进多最小支持度的概念,结合免疫遗传算法,得到改进关联规则算法;利用改进关联规则算法在用户用电数据内挖掘异常用电数据,依据排序向下闭合思想,加快异常用电数据挖掘效率;依据用户用电数据特征,得到用电量突变、同类用户差距大等用电行为险态类型,按照专家经验分析异常用电数据内的数据特征曲线和用电行为险态的匹配情况,自动化辨识用电行为险态类型。实验证明:所提模型可精准挖掘异常用电数据,具备较高的用电行为险态自动化辨识精度;在不同用户数量时,所提模型自动化辨识险态用电行为的规范化互信息值较高,具备较强的辨识鲁棒性。This paper studies the automatic identification model of power consumption risk based on improved association rule algorithm.It can accurately mine abnormal power consumption risk data and improve the identification effect of power consumption risk.By introducing the concept of multi minimum support and combining immune genetic algorithm,an improved association rule algorithm is obtained.The improved association rule algorithm is used to mine abnormal power consumption risk data in the user power consumption risk data,and the efficiency of abnormal power consumption risk data mining is accelerated based on the idea of downward closure of sorting.According to the characteristics of power consumption risk data,the types of power consumption risk,such as sudden change in power consumption risk and large gap between similar users,are obtained.According to expert experience,the data characteristic curve in the abnormal power consumption risk data and the matching of the power consumption risk are analyzed to automatically identify the types of power consumption risk.The experiment shows that the proposed model can accurately mine abnormal power consumption risk data,and has higher automatic identification accuracy of power consumption risk.When the number of users is different,the standardized mutual information value of automatic identification of power consumption risk of the proposed model is higher,and the proposed model has stronger identification robustness.

关 键 词：改进关联规则 用电行为险态 自动化辨识 多最小支持度 免疫遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]