Parameter tuning of new type energy virtual synchronous generator based on physics-informed reinforcement learning

In the backdrop of the new type power system, the prevalence of new energy generation equipment and power electronic devices is rising. However, traditional power system analysis and parameter tuning methods are inadequate for this changing scenario. To address this, the paper proposes a reinforcement learning framework based on physical information guidance to adjust virtual synchronous generator parameters. This novel approach bridges the gap between traditional methods’ applicability and effectiveness in the new power system. It also improves interpretability by combining artificial intelligence with physical models. The method’s feasibility and efficacy are verified through IEEE-39 bus system simulation experiments.