The energy consumption within urban infrastructures increases every year due to the increased utilization electrical applications. Specifically, to handle the demand and utilization in a smart infrastructure an intelligent solution is required. For this purpose, an Adaptive Edge-intelligent Green Energy Management System (AE-GEMS) is proposed in this research work which effectively combines household load optimization and electric vehicle–to–grid (V2G) scheduling. The proposed model incorporates an edge-based multi-agent reinforcement learning (MARL) for real-time control and federated learning is used to preserve the user data privacy across distributed nodes. Additionally, AE-GEMS integrates predictive load forecasting by incorporating LSTM and a policy-based V2G controller which is trained through deep Q-networks (DQN) to determine optimal energy usage, charging schedules, and energy return decisions. Also, an adaptive green score computation is proposed to evaluate the daily environmental considering the utilization ratio, peak load contribution, V2G efficiency, and carbon emission offset. The experimentation incorporates UK-DALE dataset for modeling the household consumption trends and OpenEI EV dataset for charging dynamics. The simulation analysis exhibits significant improvements of the proposed model by attaining 27.3% reduction in total energy consumption, 24.8% in cost savings, and better green score of 31.6% which is superior compared to conventional non-optimized models.
Manivannan et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: