On Predictive Control for Coordination in Multi-Carrier Energy Systems

Due to the increasing penetration of Distributed Generation (DG) technologies, such as photovoltaics, wind turbines, μCHP, biomass, etc., energy production changes from centralized units to diverse distributed units, located at lower voltage levels. In addition, installations of renewable energy resources are promoted by receiving a fixed feed-in tariff. These trends provide new opportunities for consumers, deciding when to consume or produce energy. For investigating these aspects, the energy hub modeling framework is used, wherewith multiple energy carriers, DG, energy storage systems and Renewable Energy Sources (RES) are taken into account. Energy systems are considered to consist of a number of interconnected energy hubs, which represent the interface between the consumers and the power supply infrastructures of the different energy systems. For the optimal operation of multi-carrier energy systems, a two-level control scheme is applied, which is divided into day-ahead planning and on-line dispatch. The on-line procedure adapts the pre-defined generation settings (determined by the day-ahead planner) to profile changes appearing within the day. An appropriate on-line control procedure wherewith system dynamics and forecast uncertainties can be taken into account is Model Predictive Control (MPC). By using a predictive approach, fluctuations within energy prices, load profiles and input profiles from RES can be handled in an appropriate way. In an ideal situation, the control procedure is implemented in a centralized way. In large-scale systems, however, the implementation of a centralized controller may not be possible because of practical and organizational reasons. High data transfer and large optimization problems arise. Applying distributed control, the overall problem is divided into subproblems which are solved in a coordinated way. The distributed MPC approach proposed in this