Distributed data-driven voltage control for active distribution networks with changing grid topologies.

In this paper, a novel distributed data-driven voltage control approach is proposed to handle unexpected voltage violations in active distribution networks (ADNs) with changing system topologies. In the proposed control scheme, once a voltage violation occurs, the PV inverters and battery energy storage systems will be coordinated to provide voltage corrective control. The whole control scheme is designed in a distributed and model-free manner to improve the scalability, reliability, and adaptability in regulating ADNs. In this study, the ADN is firstly divided into several local control regions (LCRs), whose models are identified online with local real-time measurements via a modified multi-layered recursive least squares (m-RLS) method. Then, based on the estimated system model, the whole control approach is realized in a distributed manner by employing a consensus version of the fast alternating direction method of the multiplier algorithm (FADMM). Simulation involving the IEEE 34-bus distribution test system verifies the effectiveness of the proposed control framework and illustrates key advantages of the proposed method, including (i) adapting to online network topology and operating-point changes without an accurate offline system model, and (ii) reaching a comparable control performance with a well-designed centralized controller. • A novel data-driven distributed control approach is proposed to regulate voltages for distribution networks with changing system topologies. • A modified multi-layered recursive least squares method is developed to achieve fast and accurate estimation of the sensitivity model for an active distribution network with topological changes. • To be consistent with the fast time-varying operation conditions in active distribution networks, the fast alternating direction method of the multiplier algorithm method is employed to coordinate the distributed energy resources. • Various case studies are conducted to demonstrate the effectiveness and advantages of the proposed control method, including identification test, control effectiveness test, and comparison with a centralized controller. [ABSTRACT FROM AUTHOR]