Dynamic Optimal Vehicle Relocation in Carshare Systems

Abstract

One-way carsharing, providing users with more flexibility on returning stations, has attracted larger market demand than has the traditional round-trip service. However, the main challenge faced by a one-way carsharing system is the vehicle stock imbalance due to the uneven distribution of user demand. This study attempted to address this problem by proposing an optimization model that integrated with a discrete choice model. The model accounts for the interdependent relationship between carshare demand and supply. User demand is influenced by the availability of carshare vehicles; meanwhile, conversely, the demand further changes vehicle availability as well as vehicle stock distribution. The model determines the optimal relocation decisions to maximize the profit for carshare operators that offer both one-way and round-trip services. This model was applied to the network of a carshare operator in Australia to evaluate the impacts of different pricing and capacity policies on the system profit. The results indicate that the one-way trip price has a more significant impact on system profit than does vehicle pod capacity. Further, the maximum profit occurs when the price of a one-way trip is around four times higher than that of the round trip.