The University of Southampton
University of Southampton Institutional Repository

Algorithms for electric vehicle scheduling in large-scale mobility-on-demand schemes

Algorithms for electric vehicle scheduling in large-scale mobility-on-demand schemes
Algorithms for electric vehicle scheduling in large-scale mobility-on-demand schemes
We study a setting where Electric Vehicles (EVs) can be hired to drive from pick-up to drop-off points in a Mobility-on-Demand (MoD) scheme. The goal of the system is, either to maximize the number of customers that are serviced, or the total EV utilization. To do so, we characterise the optimisation problem as a max-flow problem in order to determine the set of feasible trips given the available EVs at each location. We then model and solve the EV-to-trip scheduling problem offline and optimally using Mixed Integer Programming (MIP) techniques and show that the solution scales up to medium sized problems. Given this, we develop two non-optimal algorithms, namely an incremental-MIP algorithm for medium to large problems and a greedy heuristic algorithm for very large problems. Moreover, we develop a tabu search-based local search technique to further improve upon and compare against the solution of the non-optimal algorithms. We study the performance of these algorithms in settings where either battery swap or battery charge at each station is used to cope with the EVs' limited driving range. Moreover, in settings where EVs need to be scheduled online, we propose a novel algorithm that accounts for the uncertainty in future trip requests. All algorithms are empirically evaluated using real-world data of locations of shared vehicle pick-up and drop-off stations. In our experiments, we observe that when all EVs carry the same battery which is large enough for the longest trips, the greedy algorithm with battery swap with the max-flow solution as a pre-processing step, provides the optimal solution. At the same time, the greedy algorithm with battery charge is close to the optimal (97% on average) and is further improved when local search is used. When some EVs do not have a large enough battery to execute some of the longest trips, the incremental-MIP generates solutions slightly better than the greedy, while the optimal algorithm is the best but scales up to medium sized problems only. Moreover, the online algorithm is shown to be on average at least 90% of the optimal. Finally, the greedy algorithm scales to 10-times more tasks than the incremental-MIP and 1000-times more than the static MIP in reasonable time.
Electric Vehicles, Optimisation, Scheduling, Mobility on Demand, Heuristics
248-278
Rigas, Emmanouil
6f42da4c-ffea-41c0-8302-5a98c1d06a6d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
Bassiliades, Nick
81a4c6dd-bdb3-45dd-aee3-a4452b40a67c
Rigas, Emmanouil
6f42da4c-ffea-41c0-8302-5a98c1d06a6d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
Bassiliades, Nick
81a4c6dd-bdb3-45dd-aee3-a4452b40a67c

Rigas, Emmanouil, Ramchurn, Sarvapali and Bassiliades, Nick (2018) Algorithms for electric vehicle scheduling in large-scale mobility-on-demand schemes. Artificial Intelligence, 262, 248-278. (doi:10.1016/j.artint.2018.06.006).

Record type: Article

Abstract

We study a setting where Electric Vehicles (EVs) can be hired to drive from pick-up to drop-off points in a Mobility-on-Demand (MoD) scheme. The goal of the system is, either to maximize the number of customers that are serviced, or the total EV utilization. To do so, we characterise the optimisation problem as a max-flow problem in order to determine the set of feasible trips given the available EVs at each location. We then model and solve the EV-to-trip scheduling problem offline and optimally using Mixed Integer Programming (MIP) techniques and show that the solution scales up to medium sized problems. Given this, we develop two non-optimal algorithms, namely an incremental-MIP algorithm for medium to large problems and a greedy heuristic algorithm for very large problems. Moreover, we develop a tabu search-based local search technique to further improve upon and compare against the solution of the non-optimal algorithms. We study the performance of these algorithms in settings where either battery swap or battery charge at each station is used to cope with the EVs' limited driving range. Moreover, in settings where EVs need to be scheduled online, we propose a novel algorithm that accounts for the uncertainty in future trip requests. All algorithms are empirically evaluated using real-world data of locations of shared vehicle pick-up and drop-off stations. In our experiments, we observe that when all EVs carry the same battery which is large enough for the longest trips, the greedy algorithm with battery swap with the max-flow solution as a pre-processing step, provides the optimal solution. At the same time, the greedy algorithm with battery charge is close to the optimal (97% on average) and is further improved when local search is used. When some EVs do not have a large enough battery to execute some of the longest trips, the incremental-MIP generates solutions slightly better than the greedy, while the optimal algorithm is the best but scales up to medium sized problems only. Moreover, the online algorithm is shown to be on average at least 90% of the optimal. Finally, the greedy algorithm scales to 10-times more tasks than the incremental-MIP and 1000-times more than the static MIP in reasonable time.

Text AIJ-MOD-Rigas-R3 - Accepted Manuscript
Restricted to Repository staff only until 18 June 2019.
Request a copy
Archive Datasets
Download (1kB)

More information

Accepted/In Press date: 10 June 2018
e-pub ahead of print date: 18 June 2018
Published date: September 2018
Keywords: Electric Vehicles, Optimisation, Scheduling, Mobility on Demand, Heuristics

Identifiers

Local EPrints ID: 422097
URI: https://eprints.soton.ac.uk/id/eprint/422097
PURE UUID: 195263d3-3347-414b-9b21-4ac95de20a42
ORCID for Sarvapali Ramchurn: ORCID iD orcid.org/0000-0001-9686-4302

Catalogue record

Date deposited: 16 Jul 2018 16:30
Last modified: 06 Oct 2018 00:35

Export record

Altmetrics

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of https://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×