Predictive Control and Benefit Sharing in Multi-Energy Systems

This article proposes the design of a hierarchical control architecture capable of optimally coordinating multi-energy systems (MESs). A MES involves the synergetic operation of subsystems belonging to different energy domains (e.g., thermal, electrical, or gas), enhancing their energy efficiency and economic savings, at the price of significant control challenges. In fact, MESs imply an increased model complexity and the interaction of networked subsystems with largely different dynamics. This motivates the design of a multilayer control architecture where, at the upper level, a model predictive control (MPC) regulator relies on energy models of reduced order to coordinate power exchanges among MES subsystems, while, at the lower layer, decentralized MPC regulators locally control subsystems with different sampling rates, consistently with their local dynamics. On the other hand, the optimal MES regulation may imply additional costs to few subsystems, although the overall operational cost decreases. Thus, benefit sharing algorithms are also proposed, relying on game-theoretical methods, enabling to properly share the achieved economic benefit among subsystems, and guaranteeing that the MES operation is more convenient than the independent regulation for each single subsystem. The designed control strategy is tested on two different MES case studies, considering also the presence of referenced electrical and thermal networks, showing high versatility and enhanced performances.