LAUSR

This paper studies the problem of coordinated rebalancing and charging scheduling for mobility-on-demand systems with electric vehicles. A joint framework consisting of multi-server M/M/s queueing and fluid model is proposed to solve the problem, in which the former is used to deal with charging scheduling of vehicles, while the latter describes the dynamics of vehicles and users in the system. A fluid policy is presented to minimize the total number of in-transit empty vehicles under static equilibrium, yielding the optimal assignment by nonlinear programming. To cope with dynamically varying traffic conditions, we further develop a two-stage real-time policy for charging and rebalancing scheduling, where rebalancing assignment is periodically adjusted and a time-weighted averaging method is proposed to predict the future travel demand. Also, the amount of vehicles to be deployed in each charging station is given to minimize the customer waiting time for charging. The effectiveness of the proposed method is verified via simulations and experiments.