Self-Healing Q-Learning Power Routing Protocol for Smart Grids

The Internet of Energy (IoE) paradigm introduces new challenges and opportunities for efficient power routing in decentralized energy networks. A novel approach is implemented to enhance power routing efficiency within the IoE framework by integrating adaptive Q-learning techniques with a maintenance phase. The proposed distributed Self-Healing Protocol (SHP) utilizes discovered paths during the maintenance phase instead of recomputing a path to transmit a Power Packet (PP) in case of a fault. Our proposed Self-Healing Q-Learning Routing protocol aims to optimize energy delivery by dynamically adapting routing decisions and addressing system maintenance in real-time. This power routing protocol operates without centralized control, leveraging autonomous behaviors for system resilience and performance optimization. It covers all scenarios, including Multiple Source Single Load (MSSL), Multiple Source Multiple Load (MSML), and Single Source Multiple Load (SSML) scenarios. This approach is validated using MATLAB simulations.