Journées de l'optimisation 2018

HEC Montréal, Québec, Canada, 7 — 9 mai 2018

Horaire Auteurs Mon horaire

MB4 Optimization in the railway industry

7 mai 2018 15h30 – 17h10

Salle: Hélène Desmarais (48)

Présidée par Camilo Ortiz-Astorquiza

4 présentations

  • 15h30 - 15h55

    An integrated approach for inbound train split and container loading in an intermodal railway terminal

    • Bruno Bruck, prés., CIRRELT
    • Jean-François Cordeau, HEC Montréal, GERAD, CIRRELT
    • Emma Frejinger, DIRO and CIRRELT

    In the context of intermodal transportation of containers, we propose an integrated approach to a problem where one must decide how inbound trains are split into sequences of railcars, on which tracks those railcars are parked for loading and off-loading operations, and how to design proper load and block plans.

  • 15h55 - 16h20

    Collaborative optimization for train scheduling and maintenance time slots planning on conventional railway line in China

    • Changan Xu, prés., Southwest Jiaotong University
    • Shaoquan Ni, Southwest Jiaotong University

    Train services and maintenance activities of railway line are mutually exclusive, and they should be planned together in the ideal case. Nevertheless, most research have treated them as two separate planning problems. This paper proposes a collaborative optimization method for both train scheduling and maintenance time slots planning on the tactic level. Specially, through embedding the maintenance time slots planning constraints into train scheduling process, and allowing the maintenance time slots to move within pre-defined time windows. We particularly consider the minimization of the total impact of maintenance time slots planning on train scheduling. Then the problem is formally formulated as a mixed integer programming model, and a two-stage heuristic algorithm is designed to solve it. Finally, we consider Baoji-Chengdu railway line in China as a real-world application of the methodology. The results show that the CPLEX solver can efficiently obtain the approximate optimal solution within the acceptable computational time, demonstrating the effectiveness of the proposed approaches.

  • 16h20 - 16h45

    Reducing hazardous materials releases from railroad freights through hazardous materials trip plan policy

    • Omar AbuObidAlla, prés.,
    • Mingyuan Chen,
    • Satyaveer S. Chauhan, Concordia University

    This research explores the routing and scheduling of rail shipment of hazardous materials (hazmat) with blocking decisions. We consider a transportation system where different customers makes their requests for railcar moves ,i.e., both hazmat and non-hazmat freights, between different origins and destinations, with specific requirements on delivery times. we focus on minimizing total cost ,i.e., the earliness, tardiness, classification and holding costs acquired to fulfill all the demands, and the risks associated with hazmat transportations. The problem is to determine for each demand (a) the itinerary that must follow from its origin yard to its destination yard (if not outsourced), (b) the sequence of trains that it must assign along the route so that the request time, and train capacities constraints are satisfied, and (c) the blocks used to transport it for each train leg along its route. A non-linear mixed-integer programming and two MIP heuristic-based solutions are proposed for generating the trip plans. Finally, we analyze a number of medium-size problem instances.

  • 16h45 - 17h10

    The locomotive assignment problem with distributed power

    • Camilo Ortiz-Astorquiza, prés., Université de Montréal
    • Jean-François Cordeau, HEC Montréal, GERAD, CIRRELT
    • Emma Frejinger, DIRO and CIRRELT

    We introduce a variant of the Locomotive Assignment Problem which consists on determining the optimal assignment of locomotive types to trains and the choice of operation mode while satisfying power requirements. Additional constraints are included to model the requirements of the railway company. We present an IP formulation and some preliminary computational results.

Retour