Application of Reinforcement Learning for Electric Power System

Recently, with the impact of the concept of carbon neutrality turning to society, the power system once again becomes a research hotspot. At the same time, intensive learning, as a major hotspot of the development of the times, is carrying out technical linkage with more fields. This paper introduces the development of reinforcement learning, the status of modern power systems, and analyses the application of reinforcement learning in power systems, such as frequency regulation, security and stability control, automatic power generation control, and power information network. In addition, based on today's reinforcement learning as well as the development of the power system, reasonable future application scenarios are proposed, and the potential problems in the development process are also analysed. The final conclusion is that reinforcement learning has good applications in power systems, but there is still huge potential for development.