[Back]

T. Jatschka, T. Rodemann, G. Raidl:
"Distributing Battery Swapping Stations for Electric Scooters in an Urban Area";
Talk: International Conference Optimization and Applications, Moscow, Russia; 2020-09-28 - 2020-10-02; in: "Optimization and Applications", LNCS / Springer, 12422 (2020), ISBN: 978-3-030-62867-3; 150 - 165.

We investigate the problem of setting up battery swapping stations for electric scooters in an urban area from a computational optimization point of view. For the considered electric scooters batteries can be swapped quickly in a few simple steps. Depleted batteries are recharged at these swapping stations and provided again to customers once fully charged. Our goal is to identify optimal battery swapping station locations as well as to determine their capacities appropriately in order to cover a specified level of assumed demand at minimum cost. We propose a Mixed Integer Linear Programming (MILP) formulation that models the customer demand over time in a discretized fashion and also considers battery charging times. Moreover, we propose a Large Neighborhood Search (LNS) heuristic for addressing larger problem instances for which the MILP model cannot practically be solved anymore. Prototype implementations are experimentally evaluated on artificial benchmark scenarios. Moreover, we also consider an instance derived from real-world taxi and bus stop shelter data of Manhattan. With the MILP model, instances with up to 1000 potential station locations and up to 2000 origin/destination demand pairs can be solved to near optimality, while for larger instances the LNS is a highly promising choice.

http://dx.doi.org/10.1007/978-3-030-62867-3_12

https://publik.tuwien.ac.at/files/publik_293806.pdf