Graph Attention Network Based Reinforcement Learning Method for Optimal Distributed Frequency Control of an Islanded AC Microgrid

This paper proposes a new deep reinforcement learning (DRL) based data-driven method for frequency control of an islanded AC microgrid. This method aims to achieve the distributed secondary control in which each distributed generator (DG) adjusts its frequency based on the local measurements and neighboring information. For the proposed method, the control signals from secondary control loop are the nonlinear combinations of local and neighboring information, which are determined by a DRL agent. Specifically, the proposed method embeds the graph attention network (GAT) into the policy network to decide how the neighboring features are aggregated. After sufficient offline training, the weights from GAT model can be given to controllers to realize the distributed control. Simulations are conducted in a 13-bus islanded microgrid with 6 DGs, and the results show that the proposed method displays better control performance than other existing methods.