Tri-objective Stochastic Model to Optimize Location and Sizing of Scooter Battery Swapping Station

Abstract

The optimal locations and sizing of scooter Battery Swap Stations (BSSs) can promote the use of battery-swapping scooters to make cities more environmentally sustainable. Few studies investigated the optimal allocation of BSSs for scooters, especially for tri-objective optimization. Hence, this study presents the proposed Stochastic Tri-objective Grid-based Scooter BSS Allocation Model (STGSBSSAM) based on our previous studies. STGSBSSAM used ModelBuilder to semi-automatically perform data pre-processing and display the layout of the allocation of BSSs. Thus, deep learning algorithm was used to detect the flow of traffic and compensate the insufficiency of data. Also, different from our previous studies, the time slots of battery swapping demands were assumed to be highly correlated with the hourly traffic flow of different land-use types.Based on the results, STGSBSSAM could better allocate BSSs. Also, the rule of equal allocation was better than the rule of inverse distance in terms of the battery allocation rule. Meanwhile, in order to make the optimized allocation of scooters beneficial, the investigation could be made into the three objectives as a whole. The actual allocation of BSSs also indirectly validated the effectiveness of this STGSBSSAM methodology.