A Versatile Clustering Method for Electricity Consumption Pattern Analysis in Households

摘要: Analysis and modeling of electric energy demand is indispensable for power planning, operation, facility investment, urban planning. Because recent development renewable generation systems oriented households, there also a great analysing the electricity usage optimizing way to install each household. In this study, employing statistical techniques, method model daily consumption patterns in households extract small number their typical are presented. The household modeled by mixture Gaussian distributions. Then, using symmetrized generalized Kullback-Leibler divergence as distance measure distributions, extracted means hierarchical clustering. allows us capture essential similarities patterns. By experiments large-scale dataset including about 500 houses' records suburban area Japan, it shown that proposed able