Smart control systems for rural energy communities

The transition to sustainable energy systems is a critical challenge for modern society. It is particularly important in rural areas where energy poverty, land-use conflicts, and the integration of renewable energy sources (RESs) present unique obstacles. Traditional centralized energy infrastructures struggle to accommodate the variability of RESs, the electrification of transport, agriculture, and the increasing autonomy of energy users. These challenges are more critical when considering the need to balance food and energy production, ensure grid stability, and engage consumers in active energy management. Motivated by these issues, this thesis explores innovative control frameworks and data-driven methodologies to enable the reliable, efficient, and sustainable operation of future rural energy communities. The first part of this thesis focuses on increasing the RESs penetration through development of novel frameworks into smart energy systems, with a particular emphasis on dynamic agrivoltaic (AV) systems and smart parking infrastructures compatible with solar-powered electric vehicles (SPEVs). A geometric and optimization-based approach is proposed for dual-axis AV systems, enabling adjustment of solar panel orientation to maximize energy capture while respecting agricultural constraints such as crop shading and daily light requirements. Afterward, we introduced a convex model predictive control (MPC) framework for managing energy flows in parking lots compatible with both conventional electric vehicles (EVs) and SPEVs. The proposed MPC ensures operational safety, prevents simultaneous charging and discharging, and maintains feasibility even under uncertain vehicle behaviors and solar radiation, and the models are validated by real-world data. The second part of this thesis, addresses the coordination of decentralized and self-interested agents in smart control energy communities (SCECs) and energy communities (ECs). Here, a novel game-theoretic, learning-based control framework is developed, where the behavior of individual prosumers is modeled using neural networks. This approach enables distributed energy management by approximating each user's optimal response strategy and seeking equilibrium through a distributed algorithm. The effectiveness of the proposed method is demonstrated through numerical simulations, showing its potential to enhance flexibility, scalability, and user autonomy in energy communities.