| Summary: | Booming electric vehicles (EVs) lead to degraded quality-of-service (QoS) of urban public charging network (UPCN) consisting of public charging stations (PCSs) and traffic network, such as prolonged waiting time at popular PCSs. UPCN QoS analysis creates broad prospects in QoS optimization by providing reliable tools to improve UPCN performance. Conventional methods focusing on the impact of network topology on QoS underestimates effects of network capacity and its distribution among PCSs. In this paper, we proposed a heterogeneous UPCN model with limited PCS service capacity, and constructed a QoS analysis method according to PCS and EV characteristics from classic method for traffic assignment model, in which we analyzed waiting time at PCSs by queueing theory and calculated feasible routes for EVs recharging half-way by screening from deep-first-search results. Finally, we demonstrated an application of the proposed model and method in QoS optimization. We carried out experiments on popular Sioux Falls network with randomly-added PCSs. Results indicate that proportion of traditionally neglected waiting time at PCSs may reach 22% in EVs' total travel time if fast charging piles serve 2 vehicles/h, and the gap between our method and conventional method in average waiting time and spatial throughput distribution is growing as EV amount increases. These results support our method's advantages in simulating EVs' charging choice trade-off between farther routes and popular PCSs as well as higher performance in UPCN QoS optimization.
|
|---|
