Data-Model Hybrid Driven Optimal Voltage Control for AC/DC Hybrid Distribution Network

Abstract

With the development of distributed new energy and power electronics technology, the AC/DC hybrid distribution networks (AD-HDN) have gradually become an important form of distribution networks in the future and have attracted widespread attention. The high proportion of distributed new energy resource represented by distributed photovoltaic (PV) brings great challenges to the safe and effective operation of AD-HDN. Meanwhile, the randomness and fluctuation of PVs’ power output put forward higher requirements for the global coordination and dynamic response of the voltage control strategy in AD-HDN. This paper proposes a data-model hybrid-driven optimal voltage control (DMHD-OVC) method with integration of droop control and deep reinforcement learning for AD-HDN. First, parameters adjustable Q–V and P–V droop control model are established. Then, an optimal voltage control model is constructed with the aim of minimizing power losses. After that, deep reinforcement learning is employed to optimize the controllable parameters, so as to realize the online optimal voltage control for AD-HDN. Finally, numerical simulations based on a modified IEEE 33 AC/DC hybrid test system are conducted to verify the effectiveness and accuracy of the proposed method.