基于客户全视图的用电客户个性化特征挖掘系统设计

Design of Personalized Feature Mining System for Electricity Customers Based on Full View of Customers

针对用电客户个性化特征挖掘的模糊度较大、信息融合性能不佳的问题,提出基于客户全视图的用电客户个性化特征挖掘系统设计方法。提取用电客户个性化特征的关联规则分布集,分析用电客户个性化特征,采用自相关特征匹配实现用电客户个性化特征融合和聚类分析,构建用电客户个性化特征挖掘的分布式离散调度模型,采用客户全视图融合的方法实现用电客户个性化特征挖掘的寻优控制,在嵌入式的B/S构架体系下建立用电客户个性化特征挖掘系统的软件平台,建立用电客户个性化特征挖掘的客户全视图分析模型,实现用电客户个性化特征挖掘。仿真结果表明,采用该方法实现用电客户个性化特征挖掘的准确性较高,输出稳定性较好,提高了用电客户个性化特征挖掘的精度。

Aiming at the problems of large fuzziness and poor information fusion performance of power customer personalized feature mining, a design method of power customer personalized feature mining system based on customer view is proposed. Distribution set of association rules of personalized characteristics of power customers, analyzes the personalized characteristics of power customers, realizes the fusion and clustering analysis of personalized characteristics of power customers by using auto correlation feature matching, constructs a distributed discrete scheduling model for mining the personalized characteristics of power customers, and adopts the method of comprehensive view fusion of customers to realize the optimization control of mining personalized characteristics of power customers. Based on the embedded B/S framework, the software platform of power customer personalized feature mining system is established, and the customer full view analysis model of power customer personalized feature mining is established to realize the power customer personalized feature mining. The simulation results show that the accuracy and output stability of this method are higher, and the accuracy of personalized feature mining is improved.