Efficient building energy management using distributed model predictive control

• A distributed MPC approach for energy management in efficient buildings.
• Real-world operation of a solar plant for control of thermal comfort.
• Distributed MPC achieves performance comparable to centralized MPC with a suitable tuning of the parameters.

This work presents a distributed model predictive (DMPC) scheme for the efficient management of energy distribution in buildings. The energy demanded by the building's residents is supplied by a renewable power system whose capacity is limited and sometimes cannot fulfill the energy requirements of the residents, depending on the availability of renewable resources. Extensions are proposed for the distributed controllers aiming to overcome difficulties that arise from the direct application of a standard DMPC formulation. The alternative formulation retains desirable features like the ability to perform energy saving, when demand does not exceed supply, and to effectively distribute energy without disproportionally harming any of the building users, when the system experiences a shortage of energy supply. Simulation and experimental results obtained in a solar energy research center located in Almería, Spain, are reported and discussed, showing promising results for the proposed control strategy.