摘要
为了提高电费回收风险的防范能力,使用信用评价模型实现风险电力用户的识别,提升精准化营销和个性化服务的水平。基于改进蝙蝠算法、优化的适应度函数和径向基函数网络,根据电力用户信用数据的特点,构建适用于电力用户的信用评价模型,并立足于真实数据开展试点工作。结果表明,该模型可有效识别风险用户,精准推动电费回收工作的主动开展,并显著降低试点区域的欠费率,为提高效益提供技术支撑。
In order to improve the ability to prevent the risk of electricity bill recovery,a credit evaluation model is used to identify risky power users,and to improve the level of precision marketing and personalized services.Based on the improved bat algorithm,we optimize fitness function and radial basis function network.According to the characteristics of power user credit data,a credit evaluation model is constructed for power users,and application pilot work is carried out based on real data.The results show that the proposed power user credit evaluation model can improve the management and control of electricity bill recovery risks,accurately guide the development of electricity bill recovery work,and significantly reduce the arrears rates in pilot areas,providing technical support for improving benefits.
作者
薛峪峰
罗红郊
马晓琴
XUE Yufeng;LUO Hongjiao;MA Xiaoqin(Information and Communication Company of State Grid Qinghai Province Electric Power Company,Xining 810008,China)
出处
《微型电脑应用》
2023年第10期188-191,204,共5页
Microcomputer Applications
关键词
信用评价
特征选择
分类
蝙蝠算法
径向基函数
credit evaluation
feature selection
classification
bat algorithm
radial basis function