4 présentations

• 
  15h30 - 15h55

  Optimizing the preventive-maintenance plan of a public transit bus fleet

  • Charles Fleurent, prés., GIRO Inc.

  We describe a system that was implemented in the city of Angers to optimize the maintenance plan of its public transport bus fleet. Important issues related to designing an effective maintenance plan are discussed, and an algorithm is presented to generate such a plan.

• 
  15h55 - 16h20

  MIP formulations for the rapid transit line design problem for maximum demand capture

  • Souhaila Elfilali, prés., Université de Montréal
  • Bernard Gendron, Université de Montréal, CIRRELT
  • Gilbert Laporte, HEC Montréal

  The strategic problem of designing rapid transit lines for maximum demand capture consists of locating stations and segments between them to form lines, with the objective of maximizing O-D pairs coverage under topological and budget constraints. The commonly used subtour elimination constraints grow exponentially with the size of the problem, and play a key role in its complexity. The problem is consequently known to be NP-Hard.

  We propose therefore two alternative formulations using single commodity and multicommodity flow constraints which are of polynomial size. We provide the results of solving these formulations on artificial instances of different sizes (10 to 108 potential stations), randomly generated using real data from Concepcíon city in Chili.

• 
  16h20 - 16h45

  Exact solution of the evasive flow capturing problem

  • Okan Arslan, prés., GERAD, HEC Montréal
  • Gilbert Laporte, HEC Montréal
  • Ola Jabali, Politecnico di Milano

  We present a bilevel program and a branch-and-cut solution technique for the 'evasive flow capturing problem' defined as locating a set of law enforcement facilities on a road network to intercept unlawful vehicle flows who deviate from their routes to avoid any encounter with such facilities.

• 
  16h45 - 17h10

  A Markovian traffic equilibrium model for capacitated networks

  • Maëlle Zimmermann, prés., Université de Montréal
  • Emma Frejinger, DIRO and CIRRELT
  • Patrice Marcotte, Université de Montréal

  We propose a Markovian traffic equilibrium model which considers the case of networks with rigid arc capacities. This work endorses the concept of access probabilities to strictly enforce capacity constraints, playing the role of state transition probabilities in an absorbing Markov chain. We illustrate the approach on small networks.