Study On The Energy Trading Strategy Of Household Users Based On Improved Clustering And Coalition Cooperation Game

In recent years,more and more household users have access to distributed power supply,realizing the transformation of users from energy consumers to “prosumers”,which enables users to manage power generation and consumption independently and so as to achieve the optimal utility.However,due to the uncertainty of distributed energy output,such as wind and photovoltaic,and users’ electricity consumption behavior,how to effectively achieve the balance of power supply and demand of household users has become an urgent problem to be solved.Considering the controllability and flexibility of energy storage and electric vehicle charging and discharging behavior,the vehicle to home technology and energy storage system can be used to solve the problem of source-load coordinated dispatching of home micro-grid,and the peer-to-peer trade pattern can be adopted to carry out power trading among multiple users,and so as to maintain the balance of power supply and demand within the user group to the maximum extent.In addition,through clustering technology mining massive household user load data,this paper analyzes the complementary characteristics of different household users’ electricity consumption behavior,which provides the basis for the effective development of power trading between users.Therefore,this paper takes the household users as the research object,in order to solve the problems of low utilization rate of renewable energy and the imbalance of power supply and demand of household users as the foothold,and based on the improved clustering and coalition cooperation game,studies the power trading strategy of household users.The specific research contents are as follows:This paper proposes a method of extracting typical load profile of household users based on improved DBSCAN algorithm.Firstly,the input parameters of clustering algorithm are determined by nonparametric kernel density estimation method,and the neighborhood radius parameters are adaptively determined according to the distribution characteristics of different load dataset.Secondly,based on the clustering effect index,the number of columns in the nearest neighbor distance matrix which can make the clustering effect index the largest is selected as the threshold parameter of the neighborhood sample number.Finally,the input parameters of the algorithm are used to cluster the load of the household users.The simulation results show that the improved density clustering algorithm can extract the typical load profile of household users accurately.In this paper,an improved spectral clustering algorithm for load profile is proposed,which considers the comprehensive measure of similarity.Firstly,by extracting the local extremum points of typical load profile of household users,the low dimensional extremum characteristic time series is obtained,and the dimension reduction of the original high-dimensional load profile is realized.Then,the method of improved dynamic time warping is used to measure the shape similarity of load profile,and the similarity weight is calculated comprehensively by combining the distance similarity results measured by Gaussian kernel function,so as to improve the clustering effect of spectral clustering algorithm with single similarity measurement.In the case analysis,the improved algorithm is compared with a variety of clustering algorithms to verify the rationality and superiority of the proposed method in the realization of typical load profile clustering of household users.An electricity trading strategy of household users based on coalition cooperative game is proposed.Firstly,the load model of prosumer including wind power,photovoltaic,energy storage and electric vehicles is constructed.Based on this,the cooperative game model of household-household user trading price coalition is constructed and the characteristic function of cooperative game is calculated.Secondly,based on the ratio of power supply and demand,the trading price among users is improved to enhance the attractiveness of price for users to participate in power trading.Finally,the optimal strategy of satisfaction degree of sub-coalition to benefit distribution set is proposed to realize the stable income distribution of each household users in the coalition.The simulation results show that the proposed price improvement method and trading strategy can effectively motivate household users to participate in power trading in the power sharing user group composed of small-scale household users,and bring considerable benefits to consumers and prosumers.