基于线损协方差分析的群体性固定比例窃电行为检测方法

Detection Method for Group Fixed-ratio Electricity Theft Behaviors Based on Covariance Analysis of Line Loss

当一个配电台区内存在多个固定比例的窃电用户时,台区的非技术线性损耗(NTL)将由这群窃电用户的窃电量共同决定。目前根据NTL相关性强弱排序的检测方法在这一场景下有可能失效。首先,为了解决这一问题,文中发现了台区NTL和固定比例窃电用户表计数据的相关性存在着一种递增现象,并论证了这一现象成立的充分条件。以此为基础提出了基于协方差分析的窃电检测方法。通过标准化处理后的协方差来衡量NTL与用户电量之间的相关性,以求解一个组合优化问题的方式,实现对固定比例窃电用户的搜寻。然后,设计了该方法在实际应用过程中的运行框架和检测方案。最后,基于中国某省电力公司提供的实测数据和窃电模拟实验平台生成的窃电数据验证了所提方法的有效性。

When multiple fixed-ratio electricity-theft users exist in one distribution station area, the non-technical loss(NTL) of the station area will be jointly determined by the amount of electricity stolen by this group of electricity-theft users. Current detection methods ranked according to the strength of the NTL correlation are likely to fail in this scenario. First, to solve this problem, this paper finds an incremental phenomenon in the correlation between the NTL of the station area and meter data of the fixed-ratio electricity-theft users, and argues a sufficient condition for this phenomenon to hold. Based on this, a method of electricity theft detection based on covariance analysis is proposed. The correlation between NTL and users’ electricity consumption is measured by normalized covariance, and the search for an fixed-ratio electricity-theft users is realized by solving a combinatorial optimization problem. Then, the operational framework and detection scheme of the method in the practical application are designed. Finally,the effectiveness of the proposed method is verified based on the actual measurement data provided by a provincial power company in China and the electricity theft data generated by the electricity theft simulation experiment platform.