Computer Engineering and Applications ›› 2016, Vol. 52 ›› Issue (22): 253-259.

Previous Articles     Next Articles

Electricity consumer credit evaluation model based on charge payment proactivity

WANG Lian1, SUN Heping2, XIE Zhenping3, WANG Shitong3   

  1. 1.State Grid Jiangsu Electric Power Company, Nanjing 210024, China
    2.Jiangsu Suyuan High Technology Co., Ltd., Nanjing 210004, China
    3.School of Digital Media,Jiangnan University, Wuxi, Jiangsu 214122, China
  • Online:2016-11-15 Published:2016-12-02

一种基于缴费积极性的电力客户信用评估模型

王  炼1,孙和平2,谢振平3,王士同3   

  1. 1.国网江苏省电力公司,南京 210024
    2.江苏苏源高科技有限公司,南京 210004
    3.江南大学 数字媒体学院,江苏 无锡 214122

Abstract: In modern power marketing, it is a core task of reliably evaluating electricity consumers’ credit value to constantly improve the management efficiency. Existing studies mainly rely on subjective experience in designing credit evaluation models. A novel evaluation model is proposed based on consumers’ charge payment behavior in this study. In new model, a behavior index, weighted average of normalized payment period values, is firstly put forward to measure consumers’ charge payment proactivity. Then, theoretical analysis and statistical inference on real consumers’ historical payment data are performed to deduce the valid evaluation model without any known experiential evaluation rules. The application studies based on the new model also are executed, and several interesting results are reported. In a word, several different theoretical and experimental results obtained in the text indicate that, the proposed model has good reasonability and application values in quantifying consumers’ charge consumption credit values.

Key words: credit evaluation, electricity consumer, charge payment proactivity, statistical inference

摘要: 对客户进行电信用评估是电力公司实现营销精益化管理的重要任务,现有相关方法较多地依赖经验规则确定评估模型,研究提出了基于电力客户缴费积极性的信用值定量化评估新模型。首先提出以加权正则化付费周期值衡量客户缴费积极性,随后基于理论分析及客户真实缴费数据推断获得评估模型,不依赖主观经验地实现了电力客户缴费信用值的连续量化评估。理论及实验分析表明,新方法具备较好的理论依据及真实数据分析结果的支撑,能够合理有效地量化反映客户的电力消费信用。

关键词: 信用度量, 电力客户, 缴费积极性, 统计推断