A Data-Driven Approach for Smart Electricity Management in Rural Energy Communities

The current energy model, still partly based on fossil fuels and centralized systems, faces major sustainability and efficiency challenges. Transitioning to a more sustainable system requires adopting clean technologies and increasing citizen participation in energy management. Energy Communities (ECs) promote decentralized generation and shared consumption of renewable electricity. By integrating renewable sources, battery storage, and demand flexibility, ECs enable users to optimize energy consumption while actively participating in the electricity market. However, efficient resource management requires advanced models that account for dynamic electricity prices, photovoltaic (PV) generation availability, and storage capacity. This work develops a predictive modeling approach to optimize electricity consumption in a rural EC in Asturias, Spain. The objective is to minimize energy costs while maximizing resource efficiency. The predictive techniques employed range from classical statistical methods to deep learning models. A linear programming-based optimization model is formulated, incorporating PV generation, battery storage, and electricity market prices. Key constraints include grid injection limits, battery degradation costs, and storage system efficiency. Experimental results show that the optimization model significantly reduces energy costs by leveraging renewable sources and minimizing grid dependence. Additionally, intelligent energy storage control enhances the energy balance of the EC by reducing grid discharges and maximizing PV utilization.