基于数据挖掘的窃电用户预警与定位技术研究

Research on early warning and positioning technology of electricity stealing users based on data mining

为有效控制电网线损率、减轻用户窃电给配电安全性带来的不利影响、提高联合稽查工作的质量,该文针对基于数据挖掘的窃电用户预警与定位技术展开研究。根据窃电电量计算结果分析用户的实时用电行为,以此为条件设置数据挖掘权重。在此基础上,对已知的窃电样本实施过采样分类处理,分别从预警功率传输关系、定位电压传输关系两个角度,分析窃电用户预警与定位技术的应用效力。实例结果表明,应用该文技术后,各电网通路中的线损率水平均得到有效控制,证明其能够有效解决由用户窃电行为造成的配电安全问题,也为联合稽查工作的顺利开展奠定了基础。

In order to effectively control the power grid line loss rate,reduce the adverse impact of users stealing power on distribution security,and improve the quality of joint inspection,this paper focuses on the early warning and positioning technology of users stealing power based on data mining. The real-time power consumption behavior of users is analyzed according to the calculation results of stolen electricity quantity,and then the weight of data mining is set based on this condition. On this basis,the known samples of electric theft are over-sampled and classified,and the application effectiveness of early warning and positioning technology for electric theft users is analyzed from the perspectives of early warning power transmission relationship and location voltage transmission relationship respectively. The example results show that the line loss rate in each grid path can be effectively controlled after the application of the proposed technology,which proves that it can effectively solve the distribution security problems caused by the power theft behavior of users,and also lays a foundation for the smooth development of joint inspection work.