Clustering Power Consumption Data in Smart Grid

For power distributors it is very important to have detailed information about the power consumption characteristics of their customers. These information is essential to plan correctly the required amount of energy from power-plants in order to minimize the difference between the demand and supply and to optimize the load of transportation grid as well. For industrial power consumer customers, on the market the actual rate of electric power may depend on their power consumption characteristics. By using intelligent meters and analyzing their behavior, relevant information can be obtained and the consumers can be classified in order to find the best rates for the supplier as well as for the consumer. In this paper, we introduce new results on clustering the consumers. The clustering method is based on forecasting the consumption time series. The numerical results prove that the method is capable of clustering consumers with different consumption patterns with good performance as a result the forecast based method proved to be the a promising tool in real applications.