Randomized trial evaluates a new energy management system to optimize costs, carbon impact, and battery health in microgrids, suggesting a more sustainable energy solution.
Key Points
This study aims to improve energy management in residential microgrids by reducing costs, emissions, and battery degradation.
Developed an energy management system based on a closed-loop MPC framework.
Forecasting utilizes a GRU-XGBoost hybrid model for solar generation and an optimized LSTM model for load demand.
Evaluated over one week against Particle Swarm Optimization and Rule-Based Control.
Achieved a 7.1% reduction in operational costs compared to Rule-Based Control.
Saw a 12% decrease in CO₂ emissions when implemented.
Ensured smoother battery operation and enhanced resilience during adverse weather.