Reactive Voltage Optimization Control in Modern Power Systems Based on Deep Reinforcement Learning

The randomness and volatility of renewable energy and loads, along with the large-scale integration of power electronic devices in modern power systems, pose new challenges to the safe operation of the power grid. Reactive voltage optimization control is fundamental to ensuring the safe and stable operation of the grid. Addressing the reactive voltage control issues caused by the large-scale integration of renewable energy, this paper proposes a reactive voltage optimization control method for power grids based on deep reinforcement learning. A reactive power optimization model is established by comprehensively considering operational efficiency, economy, and safety. The Markov decision process is used to transform the reactive power optimization problem into a reinforcement learning sequential decision optimization problem, fully considering the temporal and spatial coupling characteristics of reactive voltage control equipment. The model is solved using the Deep Deterministic Policy Gradient (DDPG) algorithm. Finally, simulation analysis is conducted on an improved IEEE 33-bus system, and comparative results verify the effectiveness and reliability of the proposed method in the reactive power optimization decision-making process.