Multi-objective optimization for rail itineraries

The optimization of routes from an origin to a destination is a topic significantly addressed in the literature and growing due to the increasing development of public transportation. This project is the result of a collaboration between IKOS Consulting Paris and the start-up company "Trainscanner".The goal of this thesis is to propose itineraries, using the railway system, that take into account multiple parameters, representative of the trip in quantitative and qualitative terms. More specifically, in addition to the minimization of the arrival time at the destination and the number of changes, we want to maximize the customer's experience by proposing an intermediate stop along the route, dictated by the presence of attractiveness within the city. After extensive research, the Tight-Bounded McRAPTOR algorithm was chosen for this work because it appears to be efficient in optimizing problems that consider more than two parameters on large-scale public transportation networks. The algorithm was implemented in Pyhton because it is very flexible in case it wants to be adapted or integrated with other algorithms and for handling real-time data. The algorithm was tested on two databases: Trenitalia data related to Sardinia region in Italy and SNCF data related to France, Germany and Spain region.