Institute of Transportation, MOTC

       This study proposes a parking-space-flow network modeling approach to deal with the optimal scheduling problem for shared parking system considering stochastic arrival times. In this approach, uncertain vehicle arrival times are represented by discrete probability distributions. The objective is to maximize the expected profit for the shared parking system operator. The proposed model is formulated as an integer multi-commodity network flow problem and solved by optimization solver. This research aims at providing a schedule for the shared parking system operator to assign advanced parking requests to parking spaces. Computational experiments are conducted using test instances generated from a real-world shared parking example.
       To evaluate the performance of the parking space schedule obtained by solving the stochastic model, we adopt a simulation-based approach. We also compare our stochastic model with two different approaches, including the manual assignment (MA) and the deterministic model (DM). A number of scenario analyses are also conducted examine the model performance under different demand levels. The evaluation results show that the stochastic model is more robust compared to the MA and DM for generating the schedule for the shared parking system. The proposed stochastic model can be used as a decision support tool for assisting the shared parking system operator in assigning advanced parking requests to parking spaces and determining their schedules.