Smart Meter-Based Re-Phasing for Voltage Imbalance Enhancement Through Topology Reconstruction

As the deployment of stochastic loads such as electric vehicles or distributed generation has drastically increased in recent years, the need for effective distribution voltage regulation has become more prominent than ever. This paper presents a proof of concept approach for mitigating the voltage unbalance in low-voltage distribution networks by leveraging smart meters’ data. Although existing research covers methodologies to identify and/or reconstruct the grid topology, few have addressed the congestion in the distribution networks using these techniques. To overcome this gap, this paper proposes a concise approach that reduces the voltage imbalance by re-phasing consumers through correlation and regression models applied on smart meter data. The method is applied on a benchmarked low-voltage distribution network with electric vehicles, showcasing its effectiveness in relieving voltage congestion. This proof of concept demonstrates the potential of novel opportunities to preserve distribution networks against voltage imbalance by using machine learning.