Decision Analysis Of Bayesian Network In Power Marketing System Application Research

With the transformation of power supply company,power marketing is more important in the process of enterprise development and growth.In power marketing,a large amount of data is generated,and urgently needed enterprise to mine,and then provides decision support for power marketing.Because Bayesian method has unique probabilistic expression ability ? advanced knowledge learning characteristics and rich uncertainty knowledge expression,it has gradually become the focus of research and application in various fields.This study applies Bayesian network model to the power supply company marketing.Combining with related research theories at home and abroad,expounds the concept of Bayesian Network,and elaborates the process of Data Mining and the advantages of Bayesian Network Based on Data Mining(DM).At the same time,an optimized Bayesian network model is constructed by means of Bayesian network parameter learning and structural learning.Based on this theory,this study introduces the concept of power marketing,illustrates the importance of building a set of power marketing decision-making analysis system,and discusses the core content and decision-making process of power marketing decision-making,moreover the overall framework of decision analysis system is constructed.The main functional requirements include monitoring and forecasting of marketing indicators,analysis of business expansion,power market analysis,customer analysis,assistant decision-making and so on.The research focuses on the application of Bayesian network in line loss qualification rate assessment and power customer electricity risk prediction.The Bayesian network model is established by several steps,such as screening the influencing factors,determining the numerical range of Bayesian network node variables,and calculating the conditional probability table.Qualitative description and quantitative calculation are carried out to evaluate the risk of unqualified line loss and electricity consumption for selected research objects.Combined with the application of Bayesian network,line loss prediction and corresponding loss reduction measures,as well as different risk levels of power consumption prevention and other decision-making analysis based on Bayesian network model are realized.It lays a good foundation for the power supply company obtaining power marketing data?building power marketing decision support system,customer identification analysis?line fault prediction?customer fault assessment?power tariff recovery channel analysis and so on based on Bayesian network model.The results of inspection and application after decision-making show that the power marketing decision analysis based on Bayesian network model is feasible in process and reliable in result,which is helpful to the decision-making basis for the marketing work of power grid enterprises.