Applications of Reinforcement Learning in Frequency Regulation Control of New Power Systems

Abstract

With the high-proportion access of new energy, the complexity and uncertainty of new power system are increasing. The frequency stability problem becomes more and more prominent, which brings huge challenges to the operation and control of gird. Reinforcement learning (RL) is one of the most suitable methods for power system optimization and control in artificial intelligence (AI). In order to better grasp and more effectively improve RL frequency regulation control technologies, this paper reviews the research progress of RL algorithm in the field of frequency regulation control of new power systems. Firstly, the basic principle and research branch of RL are introduced. Then the applications of RL in frequency regulation control are investigated in detail for single agent RL and multi-agent RL (MARL). Finally, the future developments for applications of reinforcement learning in frequency regulation control field are summarized and prospected.