Dual-Stage Model Predictive Control Based Reduced Model Framework for Voltage Control in Active Distribution Networks

Abstract

The large-scale penetration of photovoltaic (PV) units and controllable loads such as electric vehicles (EVs) render the distribution networks prone to frequent, uncertain, and simultaneous over/under voltages. The coordinated control of devices such as on-load tap changer (OLTC), PV inverters, and EV chargers seem efficient in regulating the distribution network voltage within normal operation limits. However, the need for measuring infrastructure throughout the distribution network and communication setup to all control devices makes it practically and economically difficult. Furthermore, for large networks, the large measurement dataset of the network and distributed control resources increase the computational complexity and the response time. This paper proposes a voltage control strategy based on dual-stage model predictive control by coordinating devices such as OLTC and controllable PVs and EV charging stations. A minimum set of available control resources is identified to establish the voltage control in the network with reduced communication and minimum measuring infrastructure, using a reduced model framework. Simulations are performed on 33-bus distribution network and the modified IEEE 123-bus distribution network to validate the efficacy of the proposed control strategy.