Residential Electricity Behavior Analysis And Interaction Method Research Based On Mass Data Mining

With the smart grid data acquisition system continues to build and develop,more and more intelligent instruments are charged into the power network,to obtain real-time data of the grid.These data are not only of mass,high frequency,decentralized and other characteristics,and there is correlation between data and similarity.In the backgroud of the sale of electricity side of the reform,the potential value of these huge amounts of-data gradually mining and application.In this paper,we study the characteristics of users' electricity behavior in the smart grid.And the "network-charge" interaction model is the second research point.Firstly,the paper introduces the theoretical framework and application scenarios of the electricity grid behavior of smart grid users.Analyzes the dynamic Markov model as the user's electricity characteristics and uses the adaptive k-medoids algorithm to achieve user clustering and user's behavior analysis method.The actual massive user data is used to analyze and simulate the experiment to verify the effectiveness of the hierarchical clustering method proposed in this paper.Then,the paper studies the interactive mode of "network-chanrge" based on the analysis of user's electricity behavior,and constructs the technology of "network-load" interactive implementation.In the study of the interactive model of "net-Dutch".two types of resident users are mainly proposed:traditional electricity consumers and new residents with small power generation equipment.Then,an examplc simulation is concducteed for the interaction between two types of users and the power grid respectively.The traditional electricity consumption resident users and the power grid interactor is through the matching of the user electricity characteristic curve and the electricity price curve.And the interaction between the new generation electricity users and the grid is through the vickrey strategy.This two strategies are all effective.Finally,the paper presents a prospect for the future refinement of consumer features and the wider application of"network-charge" interactions in the future electricity market.