A Distributionally Robust Chance Constrained Model Predictive Control Based Coordinated Voltage Control of Active Distribution Networks With Uncertainties Under High Photovoltaic Penetration

This paper presents a multi-time-scale real-time coordinated control framework for voltage regulation and loss minimization in distribution networks. A distributionally robust chance-constrained control framework is presented to contain nodal power and measurement uncertainties for real-time control. Control devices, such as on-load tap changer, photovoltaic inverters, EV charging stations, and distribution static synchronous compensator, are coordinated using a two-time scale multi-step optimization model, employing receding horizon control, and leveraging real-world data sets of photovoltaics and domestic EV charging. Additionally, a local rapid response control mechanism, operating on a time scale of a few seconds, is employed to mitigate frequent voltage violations and serve as a backup measure for the centralized controller. The developed control scheme is computationally efficient and requires partial information of uncertainties, and is validated on the 33-bus network and the IEEE 123-bus distribution network. The voltage violations due to uncertainties are completely eliminated in the proposed chance-constrained solution approach, with a reduction in comprehensive voltage deviation and active power curtailment by 12–19% and 27–42%, respectively, as compared to the deterministic solution approach under uncertainties. However, there is a marginal increase in energy losses by 6–8% in the proposed approach.