Computationally Efficient Safe Reinforcement Learning for Power Systems

Daniel Tabas, Baosen Zhang

We propose a computationally efficient approach to safe reinforcement learning (RL) for frequency regulation in power systems with high levels of variable renewable energy resource…