DDPG-based local voltage active support strategy for battery energy storage systems

With the increasing penetration of renewable energy, distribution networks face growing risks of voltage fluctuations and limit violations. Traditional voltage regulation methods relying on precise mathematical models struggle to address complex distribution network structures and local information constraints. To address these challenges, this paper proposes a local voltage active support strategy for battery energy storage systems (BESS) based on the Deep Deterministic Policy Gradient (DDPG) algorithm. this paper first establishes a distribution network model incorporating BESS and formulates relevant operational constraints. Subsequently, the BESS charge/discharge decision-making process is formulated as a Markov Decision Process (MDP), with a designed local observation state space to accommodate practical scenarios with limited information availability. Finally, simulation analysis conducted on a modified IEEE 33-node test system verifies the effectiveness of the proposed strategy.