Title: Fully Distributed Peer-to-Peer Optimal Voltage Control with Minimal Model Requirements

Abstract: This paper addresses the problem of optimal voltage control using reactive power injected by inverters, aiming to keep voltages along a distribution grid feeder within a desired range. Purely local voltage control cannot solve this task under all circumstances and may even end up in detrimental control decisions, which can be avoided by incorporating situation awareness of each inverter's neighborhood. Therefore, we design a distributed control strategy with the goal of identifying the minimal amount of data and model information that need to be shared between controllers in order to solve this task. We demonstrate that short-range peer-to-peer communication and knowledge of electric distances between neighbouring controllers are sufficient. The approach was implemented and tested on a $400$~V distribution feeder with asynchronous communication channels in a synchronous optimization process, confirming its viability on real-life systems.