Design Of User Characteristic Portrait And Interactive Response Strategy Based On Electricity Consumption Behavior Analysis

With the concept of smart grid and energy Internet proposed,the traditional power industry is constantly changing,in which intelligent measuring devices are used in the power system widely,making the power system generate massive data all the time.How to make good use of power system big data resources effectively will be a problem to be solved in the power industry now and in the future.In the power system,the load of residential users often presents the characteristics of complex load feature and irregular power load curve.To realize the reasonable dispatch of the power system,it is significant to study the load feature of residential users.As a common data mining technology,clustering can extract the characteristics of users’ electricity consumption behavior and depict the characteristics of users’ electricity consumption behavior through the processing of users’ load data,so as to support the power system demand response scheduling strategy.With the access of distributed new energy and the large-scale popularization of electric vehicles,the problems of load fluctuation,widening peak valley difference and increasing peak load are brought to the power system.However,the scheduling characteristics of new energy devices and electric vehicles also provide possibility for the implementation of demand response,which means to guide users to actively participate in power system load scheduling through incentive policies,so as to achieve peak load clipping and valley filling in the power system,reduce risks caused by peak load,and maintain the stable operation of the power system.This paper proposes a demand response strategy based on the behavior characteristics of residential users’ electricity consumption.The specific work is as follows:(1)An improved K-means clustering algorithm based on Canopy algorithm pretreatment is proposed.K-means algorithm is widely used in various clustering research because of its simple and efficient characteristics.It has obvious advantages in the clustering of high-dimensional data such as daily gas consumption curve and daily load curve,but it is difficult to determine the initial clustering center and the clustering number.To solve the above problems,this paper proposes to use Canopy algorithm to pre-cluster user load data to determine the initial clustering center and clustering number of K-means algorithm.Through the example analysis,the algorithm can well solve the problem that the initial clustering center and the clustering number of the K-means algorithm are difficult to determine,while maintaining the advantages of the simple and efficient Kmeans algorithm,which verifies the effectiveness of the method.(2)An improved Canopy K-means secondary clustering algorithm based on ant colony algorithm is proposed.In the improved k-means algorithm of canopy,there are still some problems such as local optimal solution and poor clustering accuracy.By introducing the positive feedback mechanism of pheromone update in ant colony algorithm,the optimal solution can be obtained and the clustering accuracy can be improved.At the same time,the pre-clustering of Canopy K-means makes up for the problems of long clustering time and low clustering efficiency when ant colony clustering processes high-dimensional data such as daily load curve.The validity of the method is verified by a numerical example.(3)A user-side interactive response strategy based on user behavior characteristics is proposed.Based on the analysis of user behavior,TOU of peak and valley electricity price is customized to encourage users to participate in demand response.The typical load of community users is modeled,and the characteristics of electricity consumption are analyzed by the clustering method.According to the clustering results,TOU electricity prices of different categories of users are customized for different periods of peak and valley electricity prices,to achieve the goal of reducing the cost of electricity consumption and the peak-valley difference of power system load.Finally,the proposed user-side interaction response strategy based on the characteristics of electricity consumption is analyzed by a numerical example,and the effectiveness of the proposed strategy is verified.