Research On Power Marketing Inspection Of Power Supply Company Based On Clustering And Correlation Analysis

As one of the core businesses of power grid enterprises,power marketing is not only a direct source of income for power grid enterprises,but also an important communication window for power grid enterprises,power users and power generation enterprises.The quality of electric power marketing will directly affect the financial state of the company and the image of the enterprise.It is the key work of the leadership at all levels of the power grid enterprises.As an important part of marketing business,electric power marketing audit plays the role of standardizing electricity users' use of electricity and maintaining the order of normal electricity price.In recent years,information and intelligence of the State Grid Corporation to promote work vigorously,through a large number of intelligent data acquisition terminal and management system on the line,can make the marketing business departments to quickly control the electric power user state information for early warning and load scheduling and load monitoring and control is of great significance for electric anomaly detection etc..The electric anomaly detection,mainly to detect the illegal use of electricity and electric power customer behavior,the annual direct economic losses caused by the State Grid Corporation for stealing tens of millions of yuan,especially the city public user acts of stealing electricity stealing single family concealment,the loss is difficult to perceive,brings great difficulties to marketing inspection work carried out.This paper proposes the use of big data algorithm to analyze customer irregularities in electricity consumption.Taking power supply company as an example,data mining and analysis of marketing business related data is done by big data algorithm,so as to realize user's power abnormal characteristics prediction and early warning.The main contents of this paper are as follows:First,we study the basic principles and implementation process of K-means algorithm,Apriori algorithm and FP-growth algorithm.We further explore the clustering process and mapping principle of the three algorithms,and lay the theoretical foundation for the paper analysis.Secend,big data model of marketing audit,the main use of marketing related business information system data of power supply company,data mining and analysis of SCADA system line load data,95598 data,95598 complaints system repair system data,simulation experiments are carried out on the three kinds of algorithms.Third,proposed the construction of marketing information system audit system,through the study of the current marketing related business information system framework of marketing inspection related data screening finishing,direct marketing inspection data source system,thus realizing the marketing audit work of professional management.Through the introduction and simulation analysis of the three algorithms,the main advantages and disadvantages of the big data algorithms applied to the marketing inspection business are explained,so that the managers of power supply company can choose the corresponding calculation methods.At the same time,we put forward the establishment of marketing inspection system based on big data algorithm,so as to achieve real-time access and analysis of the main marketing system data,and make marketing inspection work more targeted and timeliness.