Optimizing electricity management and energy community clustering in smart grids using hybrid Physical-Neural Models

In recent years, the extensive adoption of renewable energy systems (RES) has led to an increase in electricity grid fluctuations, which can result in problems such as voltage and frequency imbalances. The unpredictability of RES, combined with the decentralization of energy production, has made the development of local energy communities a promising solution to minimize electricity losses and promote sustainable energy practices. To achieve optimal electricity management and peer-to-peer (P2P) energy sharing, a dynamic simulation model has been developed and implemented in Matlab. The model discretizes the energy community and final users into several control volumes, each of which considers different technologies and implements suitable energy balances. The model assesses energy performance for both single users and the whole energy community and investigates solutions for dispatching services for grid frequency regulation and optimizing distributed energy resource (DER) exploitation. Moreover, the model can determine the optimal clustering of the energy community users to maximize self-consumption and minimize interaction with the electricity grid. To predict the electrical load of each user, a hybrid physical and neural model, which combines an in-house building energy performance simulation tool with a feed-forward neural network, is adopted. The physical model takes into account various parameters such as weather conditions, building features, and occupancy patterns, while the neural network learns nonlinear relationships between the input parameters and the electrical load of each user. This hybrid approach improves prediction accuracy and enables the identification of optimal clustering solutions. The proposed methodology, and the related achieved results, demonstrate the effectiveness of the model in predicting the electrical load of each user and optimizing electricity management. This approach can help in promoting sustainable energy practices by enabling more efficient energy consumption and management in local communities, reducing the need for expensive grid infrastructure upgrade and ultimately leading to a more sustainable energy solution.