Robust Local Coordination Control of PV Smart Inverters With SVC and OLTC in Active Distribution Networks

Active engagement of smart inverters in grid support functions enables faster voltage regulation and increases the penetration of distributed energy resources (DERs) in active distribution networks. However, optimal control of smart inverter operations and robust coordination control of smart inverters with legacy active distribution network management are desired to fully leverage the functionality of the smart inverter. In this paper, a deep neural network (DNN)-based robust local coordination control of photovoltaic (PV) smart inverters with static VAR compensator (SVC) and on-load tap changer (OLTC) is proposed. The proposed method first performs centralized linear chance-constrained AC optimal power flow (CCACOPF) using historical data of PV output and load demand under uncertainty to obtain the robust Volt/VAR control settings of smart inverters and the optimal operation of SVC and OLTC. Then, DNNs are trained and tested as local controllers to obtain the optimal setpoints for smart inverters, SVC, and OLTC. To evaluate the performance of the proposed method, comprehensive evaluation studies were conducted on modified IEEE 33-bus systems. The results demonstrate that the proposed DNN-based local coordination control method emulates the CCACOPF-based robust coordination control method. Moreover, the performance of the proposed DNN-based local coordination control method outperforms conventional machine learning methods.