基于高级计量设施的网络窃电检测研究  

Research on Network Electricity Theft Detection Based on Advanced Measurement Facilities

作　　者：贺凌飞 He Lingfei(State Grid Xi'an Urban Power Supply Company,Xi'an 710499,China)

机构地区：[1]国网西安市区供电公司,西安710499

出　　处：《办公自动化》2025年第6期91-93,共3页Office Informatization

摘　　要：文章提出一种基于熵的窃电检测方案,通过跟踪客户消费变化的动态来检测窃电行为。相对熵用于计算从消耗变化中获得的概率分布之间的距离,当针对AMI发起窃电攻击时,消耗变化的概率分布偏离历史消耗,从而导致较大的相对熵。使用真实的智能电表数据在不同的攻击场景中测试提出的方法,结果表明,该方法检测窃电攻击的概率很高。This paper proposes an entropy-based method for detecting electricity theft,which monitors dynamic changes in customer consumption to identify theft behaviors.Relative entropy is employed to measure the distance between the probability distributions derived from consumption changes.When a power theft attack is launched against an AMI,the probability distribution of consumption changes deviates from historical patterns,leading to a higher relative entropy.The proposed method was validated using real smart meter data across various attack scenarios,demonstrating a high detection rate for electricity theft attacks.

关 键 词：高级计量设施 需求响应 负载管理 窃电检测 

分 类 号：TM73[电气工程—电力系统及自动化]