A hybrid PSO-SA algorithm for the travelling tournament problem

Sports scheduling has become an important area of applied operations research in recent years, since satisfying fans' and teams' requests and revenues of a sports league and TV networks may be affected by quality of the league schedule. While this type of scheduling problem can be solved by mathematical methods and exact solutions are accessible, it computationally leads to hard problems. The travelling tournament problem (TTP) is defined as minimising total travelling distance for all teams in a league. In this study, a new mathematical model for the TTP with the no-repeater constraint is presented. In addition, a very fast hybrid metaheuristic algorithm is proposed, which combines particle swarm optimisation (PSO) and simulated annealing (SA). Our computational experiments on standard instances show that the hybrid approach results in comparable to or even better than current best known solutions, specifically in computational time. [Received 24 August 2009; Revised 16 January 2010, 9 June 2010; Accepted 12 June 2010]