Abstract
Energy data measured on-site from buildings can help describe the consumption behavior of end-users and can be used to examine and prove certain theorems and models, that require a large volume of data to be gathered. However, the direct extraction of this data can often be a lengthily and costly process. As a result, a dataset of a residential community was constructed based on real data, where sample consumption and photovoltaic generation profiles were attributed to 50 residential households and a public building (municipal library), a total of 51 buildings. In addition, the overall power consumption of these houses was desegregated into the consumption of 10 commonly used appliances using real energy profiles. First, several consumption and photovoltaic generation profiles, as well as a vast collection of appliance profiles, were gathered. These profiles were obtained from household readings in different locations, while the public building's profile was based on the consumption and photovoltaic production profiles of the research building GECAD. The profiles went through the process of normalization and new profiles were generated to complete the number of end-users needed. Moreover, these profiles were given a maximum consumption and production level at random before being accepted by one of the end-users. Therefore, fourteen of these households and the public building were randomly attributed with renewable solar energy. Finally, if possible, the tool created allocated, at random in previously determined intervals, the appliances' load profiles into each end-user's available consumption areas.