Tune Inverter Control Gains using Reinforcement Learning

The growing complexity of modern power systems and the proliferation of inverter-based resources (IBRs) demand intelligent, adaptive, and autonomous control strategies. Traditional tuning approaches for inverter controllers often rely on linearized models, fixed assumptions, or manual parameter optimization methods that are increasingly inadequate in dynamic, data-rich grid environments. In this talk, a reinforcement learning (RL)-based framework for inverter controller tuning, presenting a paradigm shift toward data-driven, self-optimizing control, will be introduced. At the core of this approach is Control RL, an open-source reinforcement learning library purpose-built for control and power system applications. This webinar will demonstrate how Control RL enables rapid prototyping, training, and deployment of RL agents capable of achieving robust and optimal inverter control under diverse grid conditions. Beyond the theoretical foundation, attendees will see how this framework was used to develop and validate the results of our recent research on autonomous inverter tuning. By bridging the gap between RL theory and practical control engineering, this talk aims to empower researchers and practitioners to harness reinforcement learning for next-generation power system control. Co-sponsored by: University of California Riverside Speaker(s): Dr. Shuvangkar Das, Virtual: https://events.vtools.ieee.org/m/506443