Communication-Free Distributed Charging Control for Electric Vehicle Group

TL;DR

This paper introduces a decentralized, communication-free EV charging control scheme that predicts capacity margins and regulates charging behaviors probabilistically, effectively reducing overload risks without relying on communication infrastructure.

Contribution

It presents a novel communication-free, probabilistic distributed control method for EV charging, including capacity margin estimation and autonomous user decision-making.

Findings

Reduces transformer overload probability after communication failure

Accurately predicts capacity margins using GMM-based baseline estimation

Verifies effectiveness with real-world residential EV data

Abstract

The disordered charging of electric vehicles (EVs) in residential areas leads to a rapid increase of the peak load, causing transformer overload, but the charging control of EV group can effectively alleviate this phenomenon. However, existing charging control methods need reliable two-way communication infrastructure, which brings high operation costs and security risks. To offer a backup strategy for charging control of EVs after communication facilities fail, this paper proposes a communication-free charging control scheme to provide a decentralized on-site charging strategy for EV group. First, an uncontrollable EV group baseline estimation considering charging behaviors enabled by Gaussian mixture model (GMM) is proposed to acquire the capacity margin forecasting for controllable EVs. Next, this paper proposes a probabilistic distributed control method to assist users formulate the charging plan autonomously. Here, the charging behavior of EV group is regulated from an optimization with uncertain boundary conditions to a sampling with uncertain feasible regions expressed by a probability distribution. Finally, the scheme is verified via real-world EV charging data from a residential area in Hangzhou, China. The results show that this method can reduce the probability of transformer overload caused by out-of-order EV charging after a communication failure.