A metaheuristic approach to the urban transit routing problem

The urban transit routing problem (UTRP) is NP-Hard and involves devising routes for public transport systems. It is a highly complex multiply constrained problem and the evaluation of candidate route sets can prove both time consuming and challenging, with many potential solutions rejected on the grounds of infeasibility. Due to the problem difficulty, metaheuristic algorithms are highly suitable, yet the success of such methods depend heavily on: 1) the quality of the chosen representation, 2) the effectiveness of the initialization procedures and 3) the suitability of the chosen neighbourhood moves. Our paper concentrates on these three issues, and presents a framework which can be used as a starting point for solving this problem. We devise a simple model of the UTRP to evaluate candidate route sets. Finally, our approach is validated using simple hill-climbing and simulated annealing algorithms. Our simple method beats published results for Mandl's benchmark problem. In addi...

• Type: Journal
• Year: 2010
• Where: HEURISTICS
• Authors: Lang Fan, Christine L. Mumford
• Keywords: HEURISTICS 2010, Human Computer Interaction, Larger Problem Instances, Simulated Annealing, Urban Transit