ElectricICT: Low voltage loads and generators coordination towards the SmartGrid paradigm

ElectricICT is a multi-disciplinary project aimed at developing a Energy Management System (EMS) which schedules and manages home loads and dispersed generators (DG). The final goal is to improve the overall SmartGrid approach performance. ElectricICT involves scientific and technical aspects in three research area: energy, telecommunication and operations research. The project is based on a model representations of each appliance, developed starting from literature data and integrated with experimental data. An artificial intelligence approach is developed and integrated in the EMS with the aim to measure the real consumption profile of the appliances installed in the house and adapt/correct the standard mathematical model available in the database. Similarly, a detailed analysis is carried out on renewable energy resources available in the house environment, with particular attention to heat pump apparatus (cooling and heating one) and on Photovoltaic production. In particular a tool is developed to estimate, in the day ahead time frame, the Photovoltaic generation curve (taking into account information gathered on public web area: weather forecast, special events, etc.). Finally, a detailed technical analysis (on literature data) is carried out to define the energy model of the house storage apparatus (Thermal Inertia, Thermal Storage, Electric Plug-in Vehicles). The developed appliances models are exploited within the optimization core of the EMS, whose goal is to optimize the global consumption/generation power profile of the house, taking into account the citizen needs. Telecommunication solutions are applied to collect data from house appliances and to control tasks, while the user has to define the required tasks, each with the correspondent due date. The EMS core schedules the task, taking into account the overall energy consumption profile and the energy market costs. In the paper the proposed approach is presented and motivated, moreover the feature of the optimization core is detailed and discussed.