用户侧窃电检测关键技术  

Key Technologies of Electricity Theft Detection at Consumer Side

作　　者：林振智[1] 崔雪原 金伟超 刘晟源 冯小峰 冯浩洋 LIN Zhenzhi;CUI Xueyuan;JIN Weichao;LIU Shengyuan;FENG Xiaofeng;FENG Haoyang(School of Electrical Engineering,Zhejiang University,Hangzhou 310027,China;Metrology Center of Guangdong Power Grid Co.,Ltd.,Guangzhou 518049,China)

机构地区：[1]浙江大学电气工程学院,浙江省杭州市310027 [2]广东电网有限责任公司计量中心,广东省广州市518049

出　　处：《电力系统自动化》2022年第5期188-199,共12页Automation of Electric Power Systems

基　　金：国家自然科学基金资助项目(51777185);国家重点研发计划资助项目(2016YFB0900100)。

摘　　要：用户侧的窃电行为是非技术性电能损失的主要原因,并且影响配电系统的运行安全性。文中对用户侧窃电检测关键技术作了分类总结,指出了当前研究的挑战和发展方向。首先,阐述了电能计量的基本原理,介绍了随计量设备升级而新出现的常见窃电方法;然后,分析了基于电气模型与计量设备的窃电检测技术在实际工程应用中的各自优势和不足;接着,对比和总结了已有基于数据驱动的窃电检测研究中基于样本学习、异常分析和状态估计的研究方法和适用范围;最后,探讨了当前窃电检测技术在输入数据、性能评估和应用场景等方面面临的关键问题和挑战,并在窃电定位、模型泛化、策略生成和隐私保护等方面对窃电检测研究方向进行了展望。The electricity theft at consumer side is the main cause of non-technical power loss, and affects the operation safety of distribution systems. The key technologies of electricity theft detection at consumer side are classified and summarized, and the current research challenges and development trends are pointed out. First, the basic principle of electric energy measurement is described, and the emerged common electricity theft methods with the upgrading of measurement equipment are introduced. Then,the advantages and disadvantages of electricity theft detection technologies based on electrical model and metering equipment in practical engineering application are analyzed. Furthermore, the research methods and application scope based on sample learning,anomaly analysis and state estimation in the existing data-driven electricity theft detection research are compared and summarized.Finally, the key problems and challenges faced by the current electricity theft detection technologies in input data, performance evaluation and application scenarios are discussed, and the research direction of electricity theft detection is prospected in terms of electricity theft location, model generalization, policy generation and privacy protection.

关 键 词：窃电检测 智能电表 电气模型 数据驱动 异常检测 

分 类 号：TM73[电气工程—电力系统及自动化]