Q-REINFORCEMENT LEARNING BASED MULTI-AGENT BELLMANFORD ROUTING ALGORITHM FOR SMART MICROGRID COMMUNICATION NETWORK

Microgrids help to achieve power balance and energy allocation optimality for the defined load networks. The communication network in microgrids is a very complex and time-variant system that needs to reserve network resources to count on several possible situations of failure resulting in limited recovery ability and inefficient resource utilization. The network link failure can lead to imbalance network load, increased packet loss ratio, higher network recovery delay. The solution to these associated problems can be resolved by improving the Quality of Service and network reliability of the microgrid communication network. In this thesis, the focus is to enhance the intelligence of microgrid networks using a routing-oriented multi-agent system and reinforcement learning while performance assessment is carried out using network performance metrics, i.e., delay, throughput, jitter, and queue parameters.