Passenger flows, backbone and robustness in subway networks

Details

Supervisors

Delvenne, Jean-Charles

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil en mathématiques appliquées

Abstract

The importance of subway systems in urban areas is constantly increasing. Having a thorough understanding of them is essential to build and operate them properly. In this work, we present a new way of computing passenger flows through those networks. This is done by using a multilayer modelling and a new centrality measure that is based on the amount of passengers that enter each station in the network. The obtained passenger flows are then compared with the ones that are obtained without the new modelling and measure. Using the passenger flows through the network, the busiest stations and edges are identified. Lines are also compared based on the total amount of stops that are done on each one of them. The impact of modelling choices on these amounts of stops is also analyzed. Two backbones of the Paris subway network are then extracted by combining multiple sparsification filters using passenger flows as weight. The first one is composed of edges and the second one is composed of lines. This second backbone shows that the lines in the Paris subway are split into two distinct groups. A different approach on robustness based on passenger flows is then suggested that analyzes the impact of a line or station failure on the passenger flows in the network. Finally, the impacts of the line 14 extension and the deconfinement plan in the Paris subway are analyzed.