New models for the mirrored traveling tournament problem.

Abstract

The Mirrored Traveling Tournament Problem (mTTP) is a challenging combinatorial optimization problem which consists in generating a timetable for sports tournaments with two half series, what is equivalent to a double round-robin timetable problem. The distance traveled by the teams should be minimized in the final timetable, and a new objective is to minimize the longest distance traveled, named MinMaxTTP. It is proposed an integer programming formulation to the mTTP and two models with dynamic constraints to its solution. Both models are based on the detection of independent sets on conflict graphs, whose use has not been reported in the literature about the problem. Real data benchmarks from a baseball tournament are used in the experiments carried out.