Development of an approach based on multi-agent deep reinforcement learning algorithm for voltage regulation in low voltage distribution networks with high photovoltaic penetration

This paper proposes a multi-agent deep reinforcement learning based approach for distributionsystem voltage regulation with high penetration of photovoltaics (PVs). The designed agentscan learn the coordinated control strategies from historical data through the counter-training oflocal policy networks and centric critical networks. The learned strategies allow us to performonline coordinated control. Comparative results with other methods show the enhanced controlcapability of the proposed method under various conditions