Thomas Riis Stidsen, Technical University of Denmark
The bi-objective symmetric Travelling Salesman Problem (BITSP) is used as a test bed to test a Bi-Objective Branch & Cut algorithm (BOB&C). The BOB&C algorithm is implemented in call-backs, using CPLEX. The algorithm utilize the standard TSP cuts. Furthermore, a special bi-objective criteria branching approach, Pareto-branching, is utilized. Finally, the algorithm is parallelized utilizing the two-dimensional criteria space. Using this approach, BITSP with up to 300 cities are solved.