Application of deep reinforcement learning in real-time control of hybrid power flow controllers

Abstract The deep reinforcement learning techniques offers a novel approach to real-time control for hybrid power flow controllers. This paper introduces a bus voltage optimization adjustment strategy, utilizing a Markov decision process to construct a power flow control model for bus systems. A two-layer multi-agent deep reinforcement learning (MADRL) model is proposed to address the challenge of continuous action spaces in multi-agent environments. Extensive testing with large datasets has demonstrated that the proposed MADRL algorithm effectively mitigates voltage limit violations when tackling large-scale power systems, thereby providing theoretical and technical support for the intelligent development of power systems.