Novel phase transition on charging networks of electric vehicles

Abstract This study explores a novel phase transition in electric vehicle charging networks. The network is represented as an artificial traffic network populated by mobile agents representing electric vehicles. Each agent travels to a designated destination and recharges at the closest node when its battery level falls below a specified threshold. The maximum charging capacity of each node is randomly assigned via a specified probability distribution. Increasing the total number of agents is expected to raise the number unable to recharge due to the limited capacity of each charging node. Extensive numerical simulations reveal a continuous phase transition in the fraction of agents unable to recharge, characterized by the average number of charging stations per node, $ s $. At the transition point, the critical $ \bar{s} $ value for a given total number of agents, $ M $, scales as $ \bar{s}\sim M^{-\alpha} $, with the critical exponent $ \alpha $ ranging between 0.8 and 1.0.