Optimal Charging Station Location and Traffic Flow Strategy for Highway Electric Vehicles under Stochastic Conditions

The rapidly growing demand for self-driving travels by electric vehicles (EVs) during holidays has intensified the pressure on the highway network, in which EVs with restricted driving range require frequent recharges for their large-scale travel. Consequently, strategical electric vehicle charging stations (EVCS) layout is vital for EV adoption, ensuring travelers can complete their trips. To minimize the indirect environmental impact and delay caused by charging process, this paper proposes a bi-layer optimization model, in which the upper-layer model selects the new station locations under a limited number of constructions and lower layer models the equilibrium traveling flow arising from EVs choosing feasible paths under stochastic demand and charging requirement. An enhanced label-correcting technique combined with the Method of Successive Averages solves this model. Evaluated on a benchmark network, the results show that longer driving range and sufficient EVCS can both shorten the average queuing time. Specifically, increasing two new EVCS achieves a 43% reduction in average wait time. Moreover, the increase within a certain driving range can also mitigate the indirect carbon emission caused by charging. However, an excessively high driving range instead lead to the aggravation of indirect emission due to the increase in battery capacity. More stations or longer driving range can also decrease computing efficiency. These findings provide actionable guidance for highway operators to optimize EVCS planning and reveal a critical trade-off for policymakers to designate technical standards of EV, suggesting that coordinated plan can achieve both operational efficiency and environmental sustainability.