Customer Behavior Analysis In Smart Grids For Electric Power Industry

Smart grids,or intelligent electricity grids that utilize modern IT/communication/control technologies,become an international trend nowadays.Forecasting or predicting of future grid load(electricity usage)by using customer behavior is an important task to provide intelligence to the smart gird.Accurate forecasting will enable a utility provider to plan the resources and to take control actions to balance the supply and the demand of electricity.In the competitive electricity markets,forecasting of the electricity load is critical for consumers and producers of electricity for planning their operations and to maintain the risk of electricity.Forecasting also plays a very important role in economic optimization of electricity usage.In this paper,our contribution is the proposal of a new data mining scheme to analyze the customer behavior for forecasting the load of a particular consumer entity in the smart grids for a future time,which effort presents an opportunity to advance the electrical industries by addressing uncertainties surroundings questions of impact and acceptance using statically rigorous experimental methods,this paper mines the electricity behaviors of smart meter users to improve the accuracy of load forecasting,the typical day loads of users calculated separately according to different date's type(ordinary workdays,day before holidays,holidays).Second,the similarity between user electricity behaviors are mined and the user electricity loads gathered to classify the users with similar behaviors into the same clusters by using Artificial Neural Network based model i.e.the Extreme Learning Machine(ELM)which is applied to clusters to conduct load forecasting and then summed to obtain the system load.In order to prove the validity of the proposed method,we performed simulation experiments on the MATLAB platform using smart meter data from the data set provided by Shandong Electric Power Corporation China and weather data from relevant external sources in china.The experimental results show that the proposed method is able to mine the user electricity behaviors deeply,improve the accuracy of load forecasting by the reasonable clustering of users,and reveal the relationship between forecasting accuracy and cluster numbers.