Research On The User Electricity Characteristics Based On Big Data

In recent years,the rapid development of smart grid construction,at the same time the construction of a set of power users electricity information collection system,the system will provide a large amount of data for the original data analysis.However,because China has a large number of power systems,so the order of the data will reach PB or even terabyte,the traditional technical means is difficult to deal with this level of data,the need to use the emerging large data technology for related analysis.In today's social data is the most important asset,the data implied a variety of information,through the analysis of these data can be more valuable information,so that the data play the greatest role.This paper mainly analyzes the relevant electricity consumption of the user by analyzing the relevant data of the users' electricity consumption,and makes a personalized load forecasting for the user's future electricity cons umption data.First,the concept of data mining is introduced,which provides a method for data processing.In the face of massive user data,this paper introduces the common large data processing framework Hadoop and Spark,and builds a related cluster to provide a platform for data analysis for large data processing.Then,the clustering technology in data mining is applied to the large data analysis platform,and the daily load of users is clustered,and the daily load curve of the user is obtained and t he load curve is studied.Because of the high dimension of the user,the effect of the traditional clustering method is not ideal.In this paper,we obtain the iterative clustering based on the spectral clustering and apply the power iterative clustering t o the data analysis.At the same time,Platform on the relevant realization,and finally be able to get the relevant user's electricity characteristics.Finally,the load forecasting is carried out,and the local weighting algorithm combined with Hadoop is used to predict the load.Compared with the real load,the validity and applicability of the method are verified.