An autonomous electric vehicle based charging system: Matching and charging strategy

In this paper, we purpose an autonomous electric vehicle (AEVs) based charging navigation system. It accommodates charging requests by a fleet of AEVs in an integrated manner. We focus on one major problem of implementing such framework: matching between AEVs to electric vehicle charging stations (EVSEs) that minimize total operational cost. The problem is first formulated as a minimum weight perfect matching through creation of a bipartite graph that could be solved by integer programming. We further relax this to linear programming. To demonstrate effectiveness of such formulation, we implement the system with different use cases. We also propose two simple charging strategies that considers various factors, including charging price, current state of charge, and restrictions of power consumption added by utilities. Based on the system, we study their effects on the power grid and AEV owners.