Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Public Transport Network Design for Equality of Accessibility via Message Passing Neural Networks and Reinforcement Learning

Duo Wang, Maximilien Chau, Andrea Araldo

TL;DR

This work tackles the problem of designing bus lines so as to minimize the inequality in the geographical distribution of accessibility in urban regions by combining state-of-the-art Message Passing Neural Networks (MPNN) and Reinforcement Learning.

Abstract

Designing Public Transport (PT) networks able to satisfy mobility needs of people is essential to reduce the number of individual vehicles on the road, and thus pollution and congestion. Urban sustainability is thus tightly coupled to an efficient PT. Current approaches on Transport Network Design (TND) generally aim to optimize generalized cost, i.e., a unique number including operator and users' costs. Since we intend quality of PT as the capability of satisfying mobility needs, we focus instead on PT accessibility, i.e., the ease of reaching surrounding points of interest via PT. PT accessibility is generally unequally distributed in urban regions: suburbs generally suffer from poor PT accessibility, which condemns residents therein to be dependent on their private cars. We thus tackle the problem of designing bus lines so as to minimize the inequality in the geographical distribution of accessibility. We combine state-of-the-art Message Passing Neural Networks (MPNN) and Reinforcement Learning. We show the efficacy of our method against metaheuristics (classically used in TND) in a use case representing in simplified terms the city of Montreal.

Public Transport Network Design for Equality of Accessibility via Message Passing Neural Networks and Reinforcement Learning

TL;DR

This work tackles the problem of designing bus lines so as to minimize the inequality in the geographical distribution of accessibility in urban regions by combining state-of-the-art Message Passing Neural Networks (MPNN) and Reinforcement Learning.

Abstract

Designing Public Transport (PT) networks able to satisfy mobility needs of people is essential to reduce the number of individual vehicles on the road, and thus pollution and congestion. Urban sustainability is thus tightly coupled to an efficient PT. Current approaches on Transport Network Design (TND) generally aim to optimize generalized cost, i.e., a unique number including operator and users' costs. Since we intend quality of PT as the capability of satisfying mobility needs, we focus instead on PT accessibility, i.e., the ease of reaching surrounding points of interest via PT. PT accessibility is generally unequally distributed in urban regions: suburbs generally suffer from poor PT accessibility, which condemns residents therein to be dependent on their private cars. We thus tackle the problem of designing bus lines so as to minimize the inequality in the geographical distribution of accessibility. We combine state-of-the-art Message Passing Neural Networks (MPNN) and Reinforcement Learning. We show the efficacy of our method against metaheuristics (classically used in TND) in a use case representing in simplified terms the city of Montreal.

Paper Structure

This paper contains 22 sections, 10 equations, 7 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related work
  3. Mathematical programming methods
  4. Heuristics methods
  5. Graph-based Reinforcement Learning applications
  6. Model
  7. Model of the territory and of public transport
  8. Accessibility
  9. Problem definition
  10. Resolution method based on Graph Reinforcement Learning
  11. Markov Decision Process formulation
  12. High-level view of the optimization approach
  13. Message Passing Neural Network
  14. Sorting algorithm
  15. Reinforcement Learning Equality algorithm
  16. ...and 7 more sections

Figures (7)

  • Figure 1: Model of Public Transit: PT graph $\pazocal{G}$ has 2 metro lines (red points represent metro stations) and 2 bus lines (purple points represent bus stops), in addition, the blue points are the centroids and the green points are the points of interest
  • Figure 2: Accessibility example: the location on the left enjoys high accessibility as, departing from it, one can reach many PoIs in little time. On the right, instead, accessibility is poor: few PoIs are reachable and high travel times are required. The left and right locations are typical of city centers and suburbs, respectively.
  • Figure 3: Framework.
  • Figure 4: Accessibility of Montreal Metro network.
  • Figure 5: Different metric results via our Reinforcement Learning Equality algorithm against the random search algorithm.
  • ...and 2 more figures