Learning large Vehicle Routing Problems -a GNN and RL based approach

NAISS 2023/5-320


NAISS Medium Compute

Principal Investigator:

Balázs Adam Kulcsár


Chalmers tekniska högskola

Start Date:


End Date:


Primary Classification:

20105: Transport Systems and Logistics



The main focus of LEAR is on route planning for a commercial fleet of Connected Electrified Vehicles(CEVs). More specifically, the objective is to minimize energy consumption and plan charging, but also consider social aspects such as congestion, producing solutions that benefit both the fleet but also the other vehicles in the network and society in general.Extensive research efforts have been put into routing and resource allocation problems for CEVs over the years. However, there are still several aspects that can be further studied and improved. One of them is the balance between the benefits for the fleet and for the other users of the transport network. Due to the inherent complexity of the problem, the project aims at using learning methods to provide efficient solutions.