Wind farm parameter optimization identification method based on multi-agent SAC

Abstract

As more wind farms are integrated into power grid, the safe and stable operation of power system becomes increasingly challenged, making accurate wind farm modeling particularly important. Based on multi-agent soft actor critic (SAC) deep reinforcement learning (DRL), this method identifies wind farm parameters under multiple fault conditions. The method compares reactive power output curves between the detailed model and multi-agent SAC identified model, ultimately obtaining high-accuracy parameters. Finally, the effectiveness and superiority of the proposed method are verified by comparing with the identification results of Soft Actor-Critic (SAC) and Proximal Policy Optimization (PPO).