基于鲁棒深度强化学习的不完全可观测配电网逆变器电压无功控制

Abstract Inverter-based Volt-Var control plays a vital role in regulating voltage and minimizing power loss in active distribution networks (ADNs). However, a key challenge in applying deep reinforcement learning (DRL) to this task lies in the limited measurement deployment of ADNs, which leads to problems of partially observable states and unknown rewards. To address these problems, this paper proposes a robust DRL approach with a conservative critic and a surrogate reward. The conservative critic utilizes the quantile regression technology to estimate a conservative state-action value function based on the partially observable state, which helps to train a robust policy; The surrogate rewards for power loss and voltage violation are designed such that they can be calculated from the limited measurements. The proposed approach optimizes the power loss of the whole network and the voltage profile of buses with measurable voltages while indirectly improving the voltage profile of other buses. Extensive simulations verify the effectiveness of the robust DRL approach under different limited measurement conditions, even when only the active power injection of the root bus and less than 10 % of bus voltages are measurable.