2019 Competition Result

Over 50 teams participated in this year's Rail Problem Competition which seeks to optimize the Integrated train blocking and shipment path (TSBP). The list of the three finalists and their report titles and abstracts are below. The finalists will be presenting their work in October at INFORMS 2019 conference in Seattle, WA to compete for the first, second and third place.


Finalists

Title: “A New Mixed Integer Programming Model for the Integrated Train Blocking and Shipment Path Optimization Problem”

Abstract: An arc-based Mix Integer Programming (MIP) model is presented to solve the Integrated Train Blocking and Shipment Path optimization problem. No linear MIP model has been presented for this problem in the literature. We developed a linear MIP model for this problem and reported and analyzed the computational results. Three different problem instances provided by RAS2019 for this problem were solved using GUROBI optimizer.

Authors
      1. Mohammad Reisi-Nafchi, Assistant Professor, Department of Industrial & Systems Engineering, Isfahan University of Technology, Isfahan, Iran.

      2. Vahid Famildardashti, Research Assistant, CIRRELT, Montreal, Canada.

      3. Pouria Arsalani, MSc. Student, Department of Industrial & Systems Engineering, Isfahan University of Technology, Isfahan, Iran.



Title:Integrated train blocking and shipment path optimization (TBSP)"

Abstract: Viewing each OD shipment as a commodity, this problem becomes a special min-cost OD multicommodity network design problem, where the bundle constraints consider two more types of node capacities. Starting with good initial multicommodity flows, our two integer programs cover all constraints with good efficiency and effectiveness. The innovative inverse shortest path heuristic helps good estimation on the dual variables of bundle constraints. A column generation scheme is also proposed.


Authors: 
      1. TSUNG-HAN WANG, Graduate Student, National Cheng Kung University, Taiwan.

      2. YEN-TZU YUEH, Graduate Student, National Cheng Kung University, Taiwan.

      3. YEN-WEI CHEN, Graduate Student, National Cheng Kung University, Taiwan.




Title: “Integrated train blocking and shipment path optimization”

Abstract: We propose a methodology that solves the TBSPP based on three stages. First, we propose a novel formulation based on a multi-commodity network flow problem to generate an initial set of routes. Second, we greedily augment the set of routes seeking diversity in the available routes. Finally, we solve an MIP that enforces feasibility using the set of available routes. We achieve good solutions in acceptable computational time.

Authors: 

      1. Karen Angulo Díaz, Instructor, Universidad de Los Andes, Colombia.

      2. Oscar Guaje Acosta, Instructor, Universidad de Los Andes, Colombia.

      3. Daniel Suarez Bayona, Instructor, Universidad de Los Andes, Colombia.