Application of clustering methods for discovering patterns of energy use in regional areas for the residential sector

Today the wide and increasing implementation of smart meters on the household level is making it possible to better know the use of electricity by residential consumers. This contributes to improving the quality of the services provided to the consumers and also developing new strategies for planning and operation by the power grid managers. This paper analyses the discovering of patterns in the use of energy in regional areas in the residential sector, in order to propose strategies of energy management and planning at a wide regional level. In the analysis presented, both the use of electricity and the gas are considered. The energy patterns have been obtained by the combined application of two clustering techniques: hierarchical clustering and self-organized maps. The study developed took as the source of information a set of data models about the use of electricity and gas in several cities of United States. The paper presents study cases of pattern discovering in four different states and their comparison.