Research On Analysis And Application Of Residential Electricity Consumption Behavior Based On Data Mining

The electric power reform makes the new electricity selling main body enter the electric power market,and the electric power selling company gradually becomes the bridge of communication between the power plant and the users,and becomes an important part of the electric power consumption.The reform of the electricity market is constantly speeding up,so in order to acquire more potential electricity users,enhance customer stickiness and realize the stability and long-term development of the company,it is an inevitable choice to mine the massive data of users’ electricity consumption through big data-related technologies to realize personalized packages for different users.Through big data-related technologies,it is necessary to mine the massive data of users’ electricity consumption to realize personalized packages for different users.In this paper,under the background of the power-selling side reform,the clustering algorithm in data mining technology is mainly used to subdivide the power consumption pattern of residential users,and finally the effective user package recommendation service is realized.Power through the analysis of research behavior for personalized package recommendation,help to optimize the structure of residents can use,encourage users to reasonable arrangement of power load,the power service market products intelligent level,help to sell electricity company long-term stable development and management,but also to ensure the stability of the power system load,has high theoretical research value and application value.Firstly,this paper establishes the framework for analyzing the power consumption behavior of residents under the power market reform,and introduces the hierarchical flow model of the power consumption data of residents.Through data mining of the original electricity data information,the subdivision of the power consumption pattern of residents can be achieved,and theoretical and technical support can be provided for the smart grid to customize the personalized and differentiated electric power service products of residents.Secondly,the method of feature recognition and user classification is analyzed,and the user feature selection extraction method used in this paper is introduced.A distributed clustering algorithm combining DTW k-medoids algorithm and CFSFDP algorithm is proposed to process the massive household users’ electricity data set.Finally,on the basis of the first two studies,through the classification of users’ electricity behavior patterns,a recommendation method for residential users’ electricity package is proposed to recommend different packages for users with different types of electricity patterns.