Optimal dispatching of electric vehicles for providing charging on-demand service leveraging charging-on-the-move technology

Abstract

Range anxiety and charging infrastructure scarcity have been the main challenges for the mass adoption of electric vehicles (EVs). The emerging mobile electric-vehicle-to-electric-vehicle (mE2) charging technology offers a promising solution, which combines battery-to-battery and connected and autonomous vehicle technologies to enable an EV with an extra battery to charge another EV on the move. This paper focuses on the efficient pairing and routing of electricity providers (EPs) to demand (EDs) by extending the existing Charging-as-a-Service (CaaS) strategy to the mE2 charging service (referred to as CaaS+). We investigate the EP fleet management problem, which is mathematically modeled as a vehicle routing problem (i.e., mE2-VRP), aiming to optimally dispatch the minimum number of EPs to approach and serve the EDs. To adapt mE2 charging strategy to the online service involving large-scale EDs in practice, we develop a Clustering-aid Decomposition and Merging (c-DM) algorithm. It clusters the EDs according to their coalition potential so that we can decompose a large-scale mE2-VRP into smaller subproblems which can be efficiently solved by parallel computing. Our numerical experiments built upon citywide (Chicago) and statewide (Florida) case studies confirm the efficiency of the proposed c-DM algorithm. It enables us to investigate the performance of CaaS+ under a realistic large-scale setup. The results show that CaaS+ will be applied in different proportions of EV flows to save EDs’ travel time and mitigate traffic congestion to different extents in different network congestion and charging station coverage scenarios. The sensitivity analyses of EDs’ energy inventory and range anxiety also provide some hints and suggestions for improving the service efficiency of CaaS+.