A time-space network flow approach to dynamic repositioning in bicycle sharing systems

Spatial Cluster-Based Model for Static Rebalancing Bike Sharing Problem

“…The main objective of the dynamic rebalancing is to determine the optimal inventory levels that need to be maintained at each time interval. The challenges, arising in the dynamic rebalancing, necessitate the dedicated and efficient methodologies [6]. However, considering the complexity of fully integrating three key components (the user dissatisfaction estimating, the bicycle rebalancing and the vehicle routing), several partially integrated or sequential solution methodologies of the dynamic rebalancing are proposed [6], which are investigated in the following.…”

“…Zhang et al [6] proposed a non-linear time-space network flow-base formulation to fully integrate three key components of the dynamic rebalandcing (the bicycle rebalancing, the vehicle routing and user dissatisfaction). They estimated the user dissatisfaction by decomposing the forecasting period into two sub-periods based on the vehicle arrival time.…”

“…Outcomes of the model are validated for two subsequent weekdays. Analyses show that the proposed model can reduce the length of the rebalancing tour by 30%.Bike sharing systems are one of the effective and environmentally friendly solutions to the first/last mile problem; they notably reduce motorized trips, air pollution, fuel consumption and needs for parking spaces [1][2][3][4][5][6][7]. Since the last decade, these systems have become one of the main complementary modes of public transit systems in many urban areas.…”

“…This redistribution, also called rebalancing, is implemented by employing vehicles (usually trucks) to collect surplus bikes from overloaded stations and deliver to bike deficient stations [9]. The efficiency of rebalancing operations is important to the authorities to ensure that the number of bikes and vacant docks at each station are periodically restored to predefined target values [8].Rebalancing of bikes in the network is carried out in two different manners: the user-based approach, in which users can rebalance a portion of bikes in the system by their daily trips or taking bonuses for specified trips; and the operator-based approach, where bikes are rebalanced by the system operators whether in a static or dynamic manner [4,6,8,10,11]. The static rebalancing is carried out during the night when the demand in the system is insignificant.…”

“…Rebalancing of bikes in the network is carried out in two different manners: the user-based approach, in which users can rebalance a portion of bikes in the system by their daily trips or taking bonuses for specified trips; and the operator-based approach, where bikes are rebalanced by the system operators whether in a static or dynamic manner [4,6,8,10,11]. The static rebalancing is carried out during the night when the demand in the system is insignificant.…”