A MATD3 -based Voltage Control Strategy for Distribution Networks Considering Active and Reactive Power Adjustment Costs

Abstract

The rapid development of distributed renewable resources brings challenges and opportunities to the future power systems. In the article, we focus on solving one of the most important challenges – voltage control problem in a power distributed network with high penetration of photovoltaic resources. Distinguished from traditional local control, centralized control and model-based distributed control, this paper proposes a data-driven/model-based multi-agent deep reinforcement learning (MADRL) -based voltage control method while minimizing active and reactive power adjustment costs. Without the knowledge of the network topology and fully state information, the proposed method can quickly regulate the bus voltages within proper thresholds. Comparative results with alternative methods demonstrate the effectiveness of the proposed method.