基于用电行为特征大数据的异常用户识别模型研究与应用

Research and application of abnormal user identification model based on big data of electricity behavior characteristics

在落实准确执行电价的政策要求过程中,发现用户用电行为多样复杂,生成的用电数据量非常庞大,给管理增加了难度,再加上客户档案数据更新不及时,用户电价执行错误的情况一直未能杜绝。为解决电力大数据时代背景下电力工作人员难以快速识别电价执行错误的异常用户的问题,本文通过用户用电量、96点负荷曲线和多指标综合评分等手段,利用大数据算法分析目标用户群体的用电行为特征信息,构建用户用电行为特征模型,创新提出了基于用户行为特征大数据的用电异常嫌疑用户识别技术路线。本文以“福利机构用电类”用户用电行为为例,介绍了具体的模型研究思路和构建方法。该技术可以快速识别现场实际用电性质与电网公司客户档案数据不一致的情况,定位用电异常嫌疑用户。

In the process of implementing the policy requirements of accurate implementation of electricity price,it is found that the user's behavior of electricity consumption is diverse and complex,and the amount of electricity consumption data generated is very large,which increases the difficulty of management.In addition,the customer file data is not updated in time,and the situation of wrong implementation of electricity price of users has not been eradicated.To solve the staff of electric company to quickly identify abnormal user problems of electricity price error under the background of big data era,based on the user of electricity,96-point load curve and multi-index comprehensive evaluation,etc.,the big data algorithm is used to analyze the characteristic information of the power consumption behavior of the target user group,and the characteristic model of the power consumption behavior is constructed,an innovative approach based on large data of user behavior characteristics to identify abnormal power users is proposed.This paper introduces the idea and construction method of the model based on the example of“power consumption of welfare organization”.The technology can quickly identify the situation that the actual power nature of the site is not consistent with the customer file data of the power grid company,and locate the suspected abnormal power users.