Distributed Dynamic Pricing for Car-sharing Systems with Stochastic Demand Shift

This study investigated one-way car-sharing systems under dynamic pricing. Despite their convenience, one-way car-sharing systems have the limitation that vehicles can be unevenly parked according to the demand of customers. To distribute car parking, we consider introducing dynamic pricing in which customers can shift their demand (i.e., change origins and destinations by walking) according to usage prices. A model of this system is developed with consideration of stochastic processes representing human behavior in the demand shift. Furthermore, we develop a common, distributed dynamic pricing policy to minimize the unevenness of the vehicles according to the network topology representing the layout of stations. Numerical examples using a realistic traffic simulator demonstrate the effectiveness of the developed method.