Optimal Integration of Interconnected Buildings in a Smart Grid: A Bi-level Approach

Energy consumption in buildings can be efficiently reduced through energy management systems that take into account several issues like comfort, technical requirements and economic aspects. This implies a detailed schedule of plants and devices (washing machines, electrical vehicles, etc.) according to forecasting of loads and renewables. The behavior of 'active buildings' strongly affects the electrical grid and its consequent management in terms of power quality and costs. In this work, a system composed by buildings electrically interconnected is considered. Each one has a storage system, renewables, and needs to satisfy electrical and thermal demands. An architecture based on Model Predictive Control is proposed, in which, first, an upper decision maker solves an optimization problem to minimize its own costs and power losses, and provides references for power exchanges with buildings that respect power flow constraints. Then, consumers, on the basis of more detailed local information, manage storage systems and devices in order to follow the reference values, to contain costs, and to achieve comfort requirements. The architecture is applied to a case study in Genoa Municipality.