Model predictive control is a promising approach to robustly control complex energy systems, such as hybrid battery-hydrogen energy storage systems that enable seasonal storage of renewable energies. However, deriving a mathematical model of the energy system suitable for model predictive control is difficult due to the unique characteristics of each energy system component. This work introduces mixed integer linear programming models to describe the nonlinear multidimensional operational behavior of components using piecewise linear functions. Furthermore, this paper develops a new approach for deriving a strategy for seasonal storage of renewable energies using cost factors in the objective function of the optimization problem while considering degradation effects. An experimentally validated simulation model of the PHOEBUS Energy System is utilized to compare the performance of two model predictive controllers with a hysteresis band controller such as utilized for the real-world system. Furthermore, the sensitivity of the model predictive controller to the prediction horizon length and the temporal resolution is investigated. The prediction horizon was found to have the highest impact on the performance of the model predictive controller. The best-performing model predictive controller with a 14-day prediction horizon and perfect foresight increased the total energy stored at the end of the year by 18.9% while decreasing the degradation of the electrolyzer and the fuel cell.

The scheduling of hybrid energy systems with battery storage systems (BSS) and hydrogen storage systems (HSS) for the storage of renewable energies is a non-trivial task due to the nonlinear nature of electrolyzers and fuel cells and the volatile electricity generation by renewable energies. Mathematical optimization of the scheduling increases the system efficiency and decreases the share of grid electricity required to cover the electrical demand. Hence, tailor-made models are required for each hydrogen component due to the uniqueness of each hydrogen system. Therefore, the time-consuming work of model generation and validation needs to be done for every system in order to ensure adequate accuracy of the mathematical models used for model predictive control. This work derives and utilizes a simulation model of a hybrid energy system as a substitute for a real-world system. We propose a framework that uses functional mock-up units of detailed simulation models to derive tailor-made mixed-integer linear programming (MILP) formulations of the steady-state operational behavior. We combine the derived formulations for the operational behavior of each component into an optimization model of the whole hybrid energy system. The optimization model is then used for model predictive control of the simulation model. The results show that we can generate accurate models of the component behavior without detailed knowledge of the simulation model. The resulting optimization model of the whole energy system accurately reflects the simulation model and is, therefore, suitable for model predictive control.

The growing decentralization of energy systems requires scalable, flexible coordination of distributed generation, energy storage, and demand-side flexibility among local energy communities. This work builds upon the agent-based scheduling framework MASSIVE, extending its capabilities to operate in real-world settings. Within the extensive framework, agents participate in the local electricity market by submitting bids based on operational constraints and preferences of local energy components or aggregates, such as a campus. Optimized setpoints derived from market clearing are sent as control signals to physical or simulated assets. To enable the transmission to be modular, interoperable, and responsive in real time, we extend the MASSIVE framework with a lightweight, MQTT-based layer. We validate the applicability of these control signals through a series of experiments involving real hardware and technical and safety constraints. Additionally, a geographically distant battery system was incorporated in real time and it effectively followed market-driven setpoints. The results confirm that a decentralized, agent-based market coordination model facilitates flexible integration of physical energy systems. Plug-and-play functionality, heterogeneous control strategies, and interconnection across regions are collectively offered by the framework, thereby providing a robust path to smart energy communities.