基于数据挖掘技术的电力营销分析系统的设计与实现

发布时间：2018-03-09 15:00

本文选题：数据挖掘　切入点：电力营销分析系统　出处：《电子科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：国家电网公司在2006年提出“SG186”信息工程建设规划,并于2006年至2009年间,完成了包括财务、生产、营销、物资、人资等八大业务的信息系统建设。利用信息通信技术,有效地推进了电力流、业务流、信息流的三流合一,同时积累了大量的业务数据。各业务部门,尤其是售电、供电部门,如何基于形成的历史数据,发掘、利用其中的数据价值,为企业运营管理提供辅助决策,是一个急需解决的迫切问题。本文设计并实现了基于数据挖掘技术的电力营销分析系统,该系统分为两大部分的功能,包括数据挖掘建模功能及分析场景功能,数据挖掘建模功能将时间序列、分类、聚类、回归、关联规则等算法以可视化图形进行封装、集成,形成数据挖掘流组件,实现了模型的配置和建模过程的可视化;分析场景功能包括模型管理及场景应用两大模块。该系统研究了五个分析场景,包括电费回收风险场景分析、客户停电信息场景分析、用电风险场景分析、电力服务场景分析、大客户服务场景分析。整个系统中,算法研究、组件开发、封装是难点,场景建模是关键,可扩转、健壮、符合业务逻辑的模型为数据挖掘、分析工作提供了支撑。论文首先根据系统建设目标进行了需求分析工作,然后根据需求提出的目标对系统进行了设计和实现,最后对系统进行了详细的功能测试,测试结果表明系统满足业务分析的需求。本文通过构建分析场景模型,阐述了如何将数据挖掘技术应用于电网企业售电、供电业务运营管理。同时,结合不同主题所面临的业务逻辑,有针对性的选取不同数据挖掘算法,建立数据模型,得到了相应的分析结果,并利用分析结果指导业务开展,对提升售电、供电服务能力及服务品质产生了指导意义,加强了对售电、供电业务数据的企业级分析能力,使公司资源配置更优、内部控制力更强、管理效能和整体效益更高,是全面提升公司运营管理水平的重要保障。
[Abstract]:The national Power Grid Corp in 2006 proposed the "SG186" information construction plan, and in 2006 to 2009, including the completion of the financial, production, marketing, construction materials, human resources and other eight business information system. The use of information and communication technology, effectively promote the power flow, business flow, information flow of the three in one at the same time, the accumulation of a large number of business data. The business sector, especially the sale of electricity, the power supply department to historical data, based on the formation of the excavation, the value of the data, to provide decision support for enterprise operation and management, it is an urgent problem to be solved urgently. In this paper, the design and implementation of electric power marketing data analysis system based on mining technology, the system is divided into two parts of the function, including data mining model and scene analysis function, data mining modeling time series, classification, clustering, regression, association rules algorithm The visualization package, integration, the formation of data mining flow component, visualization and configuration model; analysis of scene features include model and scene management application of two modules. The system of the five analysis scenarios, including the analysis of electricity recovery risk scene analysis, customer outage information of the scene, analysis of risk of electricity scenario analysis of power service scenarios, analysis of customer service scene. The whole system, algorithm research, component development, packaging is difficult, scene modeling is the key to scalability, robust, in line with the business logic model for data mining, analysis provided support. Firstly, according to the demand analysis system construction goal then, according to the needs of the target on the design and implementation of the system, the system functional testing in detail. The test result shows that the system meets the business analysis The demand. This paper constructs the analysis scene model, expounds how to apply data mining technology to the sale of electricity grid enterprises, power supply business operation management. At the same time, the business logic with different topics facing, targeted to select different data mining algorithms, the establishment of data model, got the analysis results, and use analysis the guidance of the business, to promote the sale of electricity, the power supply service capacity and service quality have guiding significance to strengthen the sale of electricity, power supply enterprise business data analysis, make the company better allocation of resources, internal control is stronger, the management efficiency and the overall efficiency is higher, is an important guarantee to enhance the operation and management of the company the level.

【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP311.13;TP311.52

【相似文献】