<![CDATA[The growing complexity of modern distribution networks, driven by renewable energy integration, electric vehicle adoption, and dynamic consumer demand, challenges conventional centralized grid control. Smart grids, supported by the Internet of Things (IoT), provide opportunities to enhance real-time monitoring, distributed decision-making, and adaptive energy management. This research presents the design and analysis of an IoT-enabled intelligent control system for power distribution in smart grids, focusing on a layered framework that integrates sensing, communication, and intelligent control. The proposed architecture consists of three main components: IoT data acquisition from smart meters, PV inverters, and EV chargers; a control layer employing Model Predictive Control (MPC) for system-wide optimization and Multi-Agent Reinforcement Learning (MARL) for local adaptability; and a supervisory layer for visualization and utility coordination. A co-simulation environment was developed using the, incorporating renewable and demand variability, as well as realistic communication latency and packet loss conditions. Performance was assessed through comparative analysis with conventional control strategies. Results show that the IoT-enabled hybrid framework improves voltage regulation by maintaining deviations within ±5%, reduces feeder losses by 12%, lowers peak transformer loading by 18%, and decreases renewable curtailment by 22%. Furthermore, the distributed architecture demonstrated resilience against 500 ms latency and 1% packet loss, outperforming centralized MPC-only solutions. This study provides a reproducible framework for integrating IoT with intelligent control in smart grids. The findings highlight the potential of hybrid MPC–MARL systems to enhance efficiency, scalability, and resilience in distribution networks, offering practical insights for utilities and policymakers toward achieving sustainable energy management.]]>
Copyrights © 2026
